e e e e e e e e e e e e e e e e can you hear me is there microphone e e e e e e e e e e e all right everyone let us get started uh we have to get started because it's also being live streamed you do not want people to wait over there uh in the virtual space so welcome everyone uh my name is Brumar mukarji I am the chair of Michigan bio statistics I have been chair for the last six years I'm in the last hours of my chair person ship and I'm delighted to welcome you to Michigan bio statistics it h
as been my absolute privilege and honor to be the chair of this department one of the oldest Departments of biostatistics in the world uh when was Michigan biostatistics established any ideas guesses the nearest neighbor gets an award or something like that any any show hand yes please 1970 1970 next guess close not quite there 1957 1949 is the answer we are uh so you're closest nearest neighbor I'll keep you in mind for uh uh for an honor um who so I I I think that you know 1949 that means 75 y
ears of existence and this is one of the campuses where uh the statistics department is younger uh than the bio statistics Department by 20 years uh we always say that we are older and wiser uh but but many of you are probably considering statistics as well but I think that it is fantastic to see this enthusiasm for our profession uh in statistics and bio statistics so I I have this sort of like you know feeling that like you know I have to do better than gp4 so I asked like Chad GPT and gp4 tha
t like what is your advice for graduate students in biostatistics and uh the first line is very interesting as an AI language model I don't have personal experiences or emotions uh which is very important right that's where I felt like I could do better right to show you some of my personal experience and emotion but what they said as an advice is actually quite good stay motivated uh develop good study habits Network and collaborate keep up to dat with the latest research communicate effectivel
y seek our Mentor seek your mentors and take care of yourself I think these are pretty good advice as a graduate student or for a faculty uh for early career researchers these are pretty good advice so what could I tell you that is not achievable or good aable or searchable that's my again my own unique personal journey and experience so I went to grad school at produ statistics I'm not going to tell you when but it sort of like reveals that I was at a first major conference in 1999 that reveals
my age but I'm proud so uh our statistics department and it was a very theoretical statistics Department there is no data in my dissertation there is absolutely no data my advisor prohibited that and so uh uh we were in a math department building so our really dream was very altruistic there are many people who wanted to solve the reman hypothesis and so changing the mathematical Universe of over coffee and Pie that was my grad school but the grad school for me was really an window into the wor
ld in my office there was a person from Germany there is a person from Korea there is a person from United States we got to know each other's culture each other's favorite food and who we are through people different countries the window into the world the second big part and I do not recommend it was actually being a graduate student and a parent so what happened was that uh my daughter was born in the second uh year of grad school so this is truly the blackbot taken from the office of my grad
school where I spent the whole day working on some very tedious and tenacious equations and then I stepped out and someone just wiped it off and repainted a flower over it and that's really my daughter and that's really the couch in my office which we fought over the person who actually got earliest to the office got the couch so uh this is what grad school was about and my daughter was always babysat and taken care of by smart graduate students when you're poor people actually help you so there
was something really really really fascinating about being poor and like you know the proletariat not the bud so um and so I think that it was like a village a village of international students raising each other and supporting each other so these two people are phenomenal statisticians and they are in different parts of the world and I was really not a good graduate student so I'm encouraging you to embrace your failures your struggles um my thesis was uh on optimal designed for estimating the
path of a stochastic process I don't do anything remotely related to a single word of that uh dissertation so this is my advisor uh Professor bill studen so this is a before and after picture after my PhD and before my PhD Professor studen was a wonderful person he graduated with Samuel Carlin who is the father of probability and stochastic process from Stanford and got his job at p do and from 1964 to till then he retired stayed in the same job and same office so I was quite different so it wa
s a very interesting conversation between me and Professor studen but after years of toughness and like I do not have a single coauthor paper with my advisor and we can talk about that story in my personal room um but he then finally accepted me as a Facebook friend I was one of his three Facebook friends and I'm very very proud of that uh Professor studen passed away a few years ago and you know I I had the sense of being orphaned in the United States because regardless of how tough he was uh h
e was my academic parent so um what I did in graduate school I was very passionate about leadership and I was very passionate about uh communication that I should be able to communicate my technical work to a broader audience and I found two amazing female mentors and Role Models the first female chair at puru statistics Mary Ellen Bach on the leftand side and Professor Rebecca George who was the dean at carnegi melon and now the Provost at RPI and these she was an assistant professor and you kn
ow when you're are a assistant professor then graduate students are your friends because senior faculty do not pay attention to you so I and Rebecca bonded a lot and I always feel till this day every career move that I take I am the chair she's the dean so she's n plus one for me and so I always text her and say is this what I should something I should do so it's very very important to cultivate those relationships so networking is very important and I really think grad school is not about acade
mics grad school is really about friendship friendship with a group of people and you can see that we have aged some of us gracefully some of us not so much uh and so every year at uh at join statistical meetings which is our biggest conference how many of you have been to JSM yes so um so so this we get together because puru has an alumni reception just like Michigan does so you become a part of a community that's what grad school was about and that's what I think Chad GB missed because it does
n't have emotional and personal experiences so coming back from me to Anarbor what grad school is all about Anarbor is the most beautiful city in the world and you can see that this is the center which is my favorite I really like a lot of Art and uh Cinema and Michigan Theater and State Theater where I spend most of my time when I'm not looking at a computer and the sky in an oror is always blue look at it outside like it's never gray um you can quantify that so I wanted to tell you a little bi
t about the School of Public Health because many of you are probably coming from a a school of like in a in engineering or College of Liberal Arts and Sciences School of Public Health is structured slightly differently so the Departments are not physics chemistry and Mathematics the Departments are epidemiology environmental health health behavior and health education Health Management and policy so it's important for you to situate yourself in the school because many of us coming from a math St
background often think what am I doing in terms of my PS in other departments in the school but it has been really inspiring and liberating to learn from non-quantitative scientists in know into to identify important issues in public health so you can see that the we call ourselves the rocking chairs the chairs form a very Clos nit group and we have amazing leadership at the school and our Dean happens to be a bi statistician and an Alum of of the department so he understands our challenges and
that it has been a pleasure to work with the dean who is biost statistician bio statistics leadership and as I said that I grew up with seeing a lot of strong uh female mentors and I uh enjoy working in a team of with two brilliant women Kelly is here associate chair for academic Affairs and uh Professor Lu Wang and who you are going to see later on and we have a lot of fun together this discussing many different aspects of Life uh this is all of us all of the faculty as I said that this is not
just the oldest but one of the largest finest and kindest biostatistics departments and we take pride in the last word uh kindest uh we have a lot of faculty and last year we recruited nine more so we uh really really faculty are working in various different areas and mentoring students and students are big part of the research current students too many to fit and so if you zoom in zoom in it's going to just go to a microscopic level so just give let me give you some number 262 students from ar
ound the world and we are a International Community of Scholars and it has always been so people from all over the world have come to Michigan biostatistics for their training so our mission our mission um is really in terms of providing high quality training and education to the next generation of biostatistics students our graduate program is rated very very high and as you probably know that you know our department is also ranked quite high in terms of the major biostatistics department and o
ne reason is that The Graduate students and the alums uh we also consult very widely across campus bringing biostatistical design and Analysis expertise to the campus and to a wide spectrum of health related issue but being in a school of Public Health most of our work is related to health it's not it's not um uncommon or it's not rare that somebody works in a social science project or an economics project but rarely they all have some kind of connotation and implication for human health and I t
hink that our faculty and staff work on important areas of current bio statistics research developing new methodology but we also consult and collaborate widely and extensively we are committed to diversity equity and inclusion in the data science Workforce and overall we want to be and we Thrive to be an intellectually vibrant and socially Progressive Community our areas of strength and many of you have actually written these areas some of these areas um they are some of them are quite classica
l missing data longitudinal data survival analysis this is classic bread and butter bio statistics but then on top of that there are emerging areas electronic health records environmental statistics um in terms of Big Data data integration new areas are emerging and I'm not going to quite put artificial intelligence there because I do not think faculty are working as much on geni model that research is coming and I'm happy to take questions on that because I think you probably are interested so
along with our academic goals and metrics and numbers it's very very important that we walk into this room this space this department this school with some shared values and what are our shared values we definitely want to encourage bold exploration of idea not playing it safe purposeful inclusion doing a work for the broader good collaborative Spirit not just competition and self-determination and well-being these are our values and I think of in Academia we get lost in terms of goals and metri
cs and forget our values and that's something that I want to remind all of you now what does a chair do this is a chair's method so uh a chair keeps the trains running and sometimes Andor but does look like that I lied um and and and in the winter and in the summer but a chair also creates new directions and destination for the community that they belong to uh for the students for staff or faculty and in my first uh when I became chair in 2018 uh I had a vision 2020 a three-year program and I wa
nted to enhance our curriculum uh I wanted to strongly support and bolster our research Computing infrastructure I wanted the department to have a better culture integrative Department uh impactful research not just like you know published in high tier journals really translated to practice and impactful uh have a distributed form of leadership where students are stakeholders in the governance of the department and also strong biostatistics footprint uh so even you know um right now I hold a pos
ition in the office of Vice President of research and my office is very close to the university president and the provos and uh it is a unique opportunity to say that bio statistics and data science is important to the university leadership and across the university uh in postco of course during nobody wants to be cheered during covid let me tell you that but I could not predict I'm not such a good forecaster so covid happened and we persevered but postco we really came up um with a different ki
nd of vision in the late later like threeyear term the last threeyear term in terms of working really closely on our climate uh enhanced mentoring and we can talk a lot about that and the mentoring cross cutting through peer mentoring Mentor in graduate student mentoring faculty mentoring and also leadership we have a new professional development course where uh we talk a lot about leadership in professional scientific leadership Community leadership and we also have student awards that reward O
utreach and citizenship and Leadership so uh one part of the curriculum enhancement was this Masters in health data science and so we started with the concentration first and that was successful and quite in demand so we went for a full master's program the first cohort started last year with 16 students this year we received about 145 applications for this program and uh the jury is still out how many of them are going to come here so uh Big Data summer Institute this is our Flagship undergradu
ate summer research program and many of you are here I'm just loving all of you and it feels like children have come back home um this is the best thing I have ever like pursued or tried in my career and we have been running that for the last nine years we have trained 326 soldiers in data science and they are changing the world one theorem one algorithm or one and wrong code at a time uh so so uh but but you know this has been a many of you are coming from that program you know that again it's
not just about scholarship it's about Community uh I think that our department stands up out in terms of providing Computing support to the students uh we have a two full-time in-house Computing support person one uh develops R packages because many times you know you write R codes they remain in your GitHub but you don't go the extra mile to convert that into a package and a a vignette so um Mike um um Klein Sasser is our our package and he's a wonderful person and then uh Jacob is coming and D
an Barker was our cluster support person but he left for a fantastic job uh but Jacob is coming from next Monday U really assist people with high performance Computing and cluster Computing so we have a department software page and I strongly encourage you to actually go and go to the departmental web page and bio statistics Computing resources software and you can see the incredible amount of software and tools faculty are developing you you can search by area you can search by faculty this is
our product right like if people ask you that what do you do which is democratized which is going to be used by others so software is one currency which we know because technical papers are after all for a limited audience so it's very important if you develop a method to translate that to the software and we have that commitment and we put resources to it this is something is very important for a department that if you talk about something then you have to enable the community to pursue that pa
th and so I think that's very important we also have a departmental inclusion and wellness Advocate uh Chima is working with our students uh organizing um events so that they're mindful about their Wellness about stress busting events uh this is a joint position with epidemiology so she also organizes a lot of joint events with epidemiology so what is our goal we want to produce brilliant scholars in statistics and Bri statistic but it doesn't really matter uh if you're not happy in the end so w
e really want to create a foster a community so that we have happier graduates of tomorrow and our students are amazing you'll meet the current students they're the best part of A Faculty life uh they organize many many events some intellectual and some just pure fun uh graduate student seminar Journal club brownbag uh biostatistics Student Association statcom this is our Flagship organization working with nonprofit organizations Dr spino is here who has been involved in the leadership of statco
m for many years and also events on diversity equity and inclusion so I'll just give you one project that I have now recently undertaken it's actually to uh showcase and catalog the careers of women who have taken research leadership role which who are former graduates of the department in the last five years 5 to 10 years actually and you can see the different faces of leadership and our students have been really really remarkable in proving themselves in the outer world and I just want to show
you some of last year's graduates career so Fatima graduated last year class of 2023 I just picked three uh from this style of brilliant and powerful and empowering women uh is an assistant professor of biostatistics in Boston University inma is an assistant professor of biostatistics in Brown University and they're all the last class Margaret Banker assistant professor in North un Northwestern University so I think that uh sometimes we just struggle with work life balance and a lot of question
s but these Role Models I want to showcase their kids their uh careers they are doing leadership in Academia government and Industry and this is U an opportunity for a current and incoming graduate students to know know about their career so uh we have a lot of community building events uh lot of fun uh we are big enough so that we can actually rent a movie theater and fill it up so last uh month we watched Oppenheimer together and students vote on the movie um and so the last year the previous
year it was um everything everywhere all at once that like really jarring movie and so um and so I i i i and we have so many events we have you know Pie Contest I'm going to show you something so last uh March 14th on the Pi Day we had this enormous spread of Pi and our staff are incredible you have probably figured that out and then we had a countdown to like at 313 and you can see that the 3 2 1 so this was our pie drop so so if you did not figure it out it's just like the Pug drop the pie was
dropped and uh one student could remember 160 places of decimal he got a special award so it's all the kind of like and that's deage right here right here and his brother is SW and uh they look similar so um they will be here today and so congratulate him for remembering 160 places of decimal of the pie uh so but but so this is I don't know I I don't want to do that again okay so uh so the so as I said that you know uh students really like we go for the for the new students we have a retreat in
cam store uh true camping um feeling uh when I came to the United States I did not know what s'mores were till like after 15 years of my immigrant life uh so I wanted everyone who is coming to know what s'mores are on their first week so so that's and this is a discovery which should not wait so long in your life so I I I really wanted that to happen we also have a wonderful lecture uh sequence called Journey lecture we had a brilliant lecture by Dr Kidwell yesterday and it was so amazing and y
ou can see if you wanted to see a lot of baby pictures this is your moment and this has been a really galvanizing moment for the community because we only see people through our CVS we do not know their childhood what they experienced their struggles how they fell in love uh and so all of these secrets are revealed in a journey lecture so uh I I I really love it I think that you can go to many other departments you are not going to find the department which is so good in terms of food party and
skits so we are good at skits and this is you can see me our former chair and a like a seminal researcher Rod little and Dr kidal and we're going to show you and this is on our YouTube channel I really want you to subscribe to our YouTube channel we have very few followers and I I really even if you don't come here please please do subscribe to our YouTube channel so this is about vaccines but I'm not going to show you uh I'm going to spare you please do that like in your own time but this is a
really uh a skit about like you know talk about vaccines and so really serious topic but packaged in a humorous way and uh and Kelly was worried about vaccinating her children in the skit in the skit in the skit in the skit in the skit so so so uh when covid came you know uh so covid-19 was a time where all of us were thinking how to contribute to the society how to contribute to this crisis and so um our our collaborators and our colleagues in the medical school were working Round the Clock in
um in emergency rooms and in ICU saving lives living in their basement living in their garages and so faculty in biostat really um Rose to location and there are so many faculty in involved in so many impactful projects so you can see that um and there's a lot of collaboration within the department for example Dr song's group uh proposed this very nice cool package and uh methodology called esir for modeling the pandemic in uhan and then our team actually borrowed that tool and wrote a very time
ly paper which really was influential in governing India's uh Public Health policy so this was our paper very quickly and we are all working in a short quick time around turnaround time uh we also looked at not just India or China but really here in our own academic medical system where the uh outcomes the covid outcomes were similar across racial ethnic subgroups so sometimes I think that time comes when you have to forget everything and there is no choice except to work on some problems and co
vid was one of them and this was also a very important paper by led by one of our then graduate students about how much de delay in screening and cancer treatment uh how much that is changing the survival estimates for cancer patients and this has been used really really widely these are highly cited papers and led to so many media mentions and became the department became very important in terms of the national eye in terms of the quantitative work that we are doing for covid so I think I have
sold the department to you but again if you are still um conflicted you're thinking whether to come here or not then this song is going to change your mind so uh no other department has an Anthem like this so I'm going to play and Rod is a very talented singer I'm not going to sing but my name is Rod little and I've been in bi statistics since 199 3 when I came as the chair I love the department so much that I even made up a song about it here's the first verse we are biostat Michigan's where we
're at and we love all those symbols and cfts others think our te tests are dull and flat but our data are barrel of laughs we are biat Michigan's where we're at with a SASS and a PC3 should old equations be forgot biostat is the place for me so I uh just like uh Rod said that biostat is the place for me I certainly have found my place in this department for the last 18 years and I hope many of you find your place in this department but most importantly in the broader field of biostatistics I ca
nnot tell you how enthusiastic and I and inspired I am for our field and so I hope you join Michigan bio statistics but I'm so thrilled that you are interested in BIO statistics that's so important to me i' be happy to take any questions that you have about the program about the future about uh statistics about careers anything do you want me to do something oh oh oh oh I can throw them yes Kelly is basketball player so she will throw them thank you come on I love this microphone so you showed u
s your preco kind of plan your postco plan do you have an idea of where you want to see the department go in the next five years so um I think that you know as I'm transitioning from the chair position every chair has their own sort of vision right so I think that but the vision is actually built by input from faculty and students and staff from their community so um the next chair will Define the vision but I definitely think that if I as a senior faculty can speak to that uh and also our commu
nity of Faculty I think we need to be a little bit Bolder in terms of our integration with computer science I think that would be very important for us as artificial intelligence and these new models are really becoming very powerful we need to be a voice in the game and the second thing uh we have been thinking a lot about this that how to strengthen our undergraduate offerings right so that's also very important to us the school has a undergraduate in public health and there are some courses i
n biostatistics but we are actually launching more courses so that should be one part of the agenda we also I think that we have a long way to go one of my marked failures have been to recruit uh students faculty and staff from truly diverse backgrounds and I think that it has to be a priority it's not easy but to work from grassroot Level towards building the pipeline to really recruiting and fostering a cohort of truly diverse Community I think that has to be one of the priorities these will b
e uh definitely priorities uh to maintain the size right so we have grown a lot and with every growth comes some pain and I think that together we have fought really hard that after the growth pain we all stand taller not as tall as her but still like you know taller uh so I I I do think that every Community goes through this growth pains this expansion pains and how to really navigate that and still provide worldclass education for example you know as we grew we have two sections of every each
of our core courses so the classrooms are not so big we have tutoring for the firste students we have um really tried to incorporate peer mentoring so that the students learn from each other uh we have modified the qualifying examination so I think that a department in agile and active if it's sort of transforms with the change so but I would think that these will be definitely uh priorities thank but you know our field is changing so fast when 2022 November all the large language models were an
nounced people did not foresee that what would happen there would be a job title called prompt engineer right I could not really talk about that and then there is a course on prompt Engineering in Michigan School of information so our field is changing very fast so we have to be agile and you have to teach students to have those skill sets so that they can pivot to new areas that I cannot predict right now yes please um yes I have a question this is so weird do you want to uh say your name and a
s well yeah my name is Eva um I was wondering if your department um collaborates with other departments in the university for research and that kind of thing yes so we collaborate really really widely and so Michigan is uh University of Michigan and that's something to really think about because there are only very few universities who just so strong cross cutting in medicine and Public Health in Pharmacy in social sciences it's really iconic um that ISR which is the one of the oldest survey org
anizations in the country and actually leads some of the nation's biggest studies is actually in Michigan so how many of you have used the Lut scale in terms of like you know putting serving numbers and so on and so uh yeah so Lut was actually a faculty in ISR and so it's it's really a very place of great history and um so with the statistics Department with the biostatistics Department bio informatics this is the quantitative Sciences are very strong but the clinical Sciences are very strong an
d so our main uh clinical collaborations in the department actually lie in um cancer in kidney dis where we have different centers in terms of coordinating data coordinating centers in terms of statistical genetics we probably collaborate with almost every department on campus but mostly in health uh and there are crosscutting institutions that you should look up and three of them are Institute of social research where many of our faculty are engaged Institute of Health policy and Innovation whe
re some of the work that we do is gets translated to policy as well as Midas which is the miss Michigan Institute of data science faculty are really integrated into those uh but also in cancer you know in the cancer center leadership I'm one of the uh associate directors for the cancer center uh in cardiovascular diseases in pain diseases so we have really strong footprint in the biomedical and public health but within Public Health also we have a lot of collaboration with epidemiology and envir
onmental health where the data are getting more complex and so uh to answer your question that there is a lot of work and you'll see from mikel's uh presentation as well to really think about climate change and environmental modeling and spao temporal modeling um and how does it relate to Downstream outcomes like agricultural outcomes and health outcomes that's also a big area which is coming up where we are collaborating with many schools yeah thank you hello strange this St um my name is T and
I'd like to ask um about your invasions for international students in the department so uh as I mentioned you probably saw in my presentation that uh I really believe in International Scholarship I came to the us as an international student in 1996 and I did not know anybody right like I do not have any friends I I I do not have any family uh I feel that the academic Community really was my home and helped raise me as a human right so that's how we envision it um and the moment you leave your c
ulture your language your food and step into an unknown territory it's a courageous move and in that moment of Courage you also see that your exploration is wide open and this and and I I strongly encourage and I benefited a lot by stepping outside my comfort zone and making friends from different cultures and really knowing about the world and so we believe in this community that we treat everybody with respect but we also try to learn about our cultures so I'll give you a couple of examples on
e is that we just celebrated our Lunar New Year a dumpling making event where everybody participates we celebrate Diwali we celebrate Lunar New Year we have Multicultural game nights because different countries have different games and you know in a bi Department people love dreams so you can you can see so that would be very attractive and these Journey lectures also showcases careers and lives of people and we serve food from the country or the place of origin so yesterday we had special Maryl
and granola uh because she is born in Maryland so I I I I just think that to embrace all cultures and treat everybody with respect is very important to us um and I think that there is a lot of adjustment right cultural adjustment social adjustment when you uh come to a new country and also get integrated with a new culture uh there is a assimilations there is social theory that you can belong to assimilation separation or integration three types of environment and one of the way that chidimma wo
rks with the students is to really understand one oneone what their struggles are so I would say that and also your uh seniors form a very welcoming community in terms of the existing grad uate students so I really think that we have a lot of work to do in integrating the different cultures in a seamless way in the department but uh I think that I think that this is a global community of Scholars and you can see that reflected in our faculty you can see that reflected in our students and I'd als
o like to mention our summer program is one of the very few summer programs which uh uh accepts International applications because we have a separate um separate like fun through a donor which actually that's a priority for us because usually training grants are restricted to US citizens and permanent residents so this has always been a priority our commitment to our students and to an individual other questions you're feeling good about graduate school excited yes scared scared why tell me why
you scared yes because I saw like a like a very appreciable nod so tell me why you are excited and why you are scared maybe everybody shares that same hey everyone I'm Alex um I'm excited because for one I want to move out of Texas that's a big but I think it's the next step for me I love the next step I'm nervous I think just because it it sounds hard and that's I mean I love a challenge yes but I want I want to make sure I feel supported in the place I go to and I'm going to be leaving a lot o
f people behind yes and I mean leaving Texas behind don't get me wrong but also leaving you know friends and family that's kind of scary to me yeah so didn't expect to be vulnerable but here we are so this is the first step right like this is a say this is a space which I think is safe and brave and that's that's very important that we are we do not always have to look Invincible right we we have our vulnerabilities we leave our Roots behind in order to pursue something bigger in life and we can
talk about International students but all of us many very few of us have family nearby we leave something to gain something and to gain that insight and there are many places where actually graduate school or 10e faculty life you'll say oh this is very hard let me see whether you can succeed you have to be smart but I do think that I can honestly vouch for this that the moment I entered Michigan I was a very like you know young parent and completely lost like a failure of a academic at the time
um things change because uh I was going through a difficult time in life adjusting to different forces that life is about work is not just life and I struggled in the beginning but nobody made me feel that oh let us see whether you can succeed it was everybody trying to help me and support me to figure out a way how to make me succeed that is extremely important that where people are not always judging you for your Brilliance but really supporting you to be your best and I think that was very i
mportant to me and I really felt that in Michigan and it's okay to be nervous it's I I I had a Italian Friend um and he will always say that it's okay to cry so I think that it's okay to cry it's okay to be nervous it's okay to be afraid afraid of failures that's only through which facing our fears are the way we grow but there is a bigger pursuit of doing something really really important with the talents that you have making a difference and see a room full of friends and faculty everybody is
there to support you elevate you question Dear Professor is it because we are a thank you dear Professor it is it because we are a more combined subject it's harder for AI to take place than pure statistics can you share something more with our advantages or threats to AI developing so uh this is a great question so let me give you some perspective one is that I I do think that there will be tremendous application and integration of AI into biostatistics um I do think that uh so far it has been
mainly the methods have been coming from computer science and engineering schools but uh I think it's soon going to see massive integration with bio statistics I'm part of the university-wide AI resource committee as well as um the assistant vice president for research data strategy and we are seeing a lot of effort I'll just give you one example these AI tools and large language models are really really good in analyzing data which are multimodal for example um text me text messages or like you
know clinical notes uh voice messages images and all of the structured data and the unstructured data together so for example in analyzing electronic health record once we can create a environment where person protected health information can be securely analyzed then we are going to see massive massive influence in terms of data integration and Analysis of heterogeneous multimodal data because they just have that Foundation models have that flexibility of really taking different data types if
you say I'm going to do a statistical model which has images as an object or uh it has voice messages or clinical notes as an object it's sort of UN in so it's not really naturally intuitive for statistical models to think like that but I think that there are many models and AI models I and also we need to understand the properties of this model evaluate this model but I also share a lot of concern in terms of fairness and ethics of these AI models if you think about the current AI models their
training data is coming from a very selected group of people in the world and so if you're building your model on a very biased training data and policies and decisions are being made on that biased data and biased algorithm then inequities are going to really reinforce themselves and we need representation in terms of the training data and then before we apply to test data all over the world or validation data all over the world so design how to design and evaluate these AI tools statisticians
are going to play a humongous a role in that and we are seeing a lot of conversations this has already like you know taking uh different professional learned societies attention and we are building forces to understand uh we may be a little slow statisticians are usually very deeply rigorous people but I think that we are going to make tremendous contribution in innovation in terms of design and evaluation of these tools and the other thing I wanted to mention that uh there are areas in which un
iversities can really use these AI tools that the areas which may have Global and positive social impact so uh for example you know uh the application of AI to climate and sustainability the application of AI to education the application of AI to human health application of AI to Poverty Solutions these are areas where I think that our existing strength can blend where probably uh big corporations and industries are not going to invest as many resources so AI with human centered Ai and AI for po
sitive Global impact is where I think that universities are going to make tremendous contribution you're good all right thank you so much [Applause] I have a wonderful weekend here so that's very important I'm going to stay for your talk from aell perzi one of our newest faculty members um although he's been here a year but uh we're so grateful to have him here today to Talk N months only with us I apologize I was not using the microphone um and so while he's bringing up slides um where so we he
ard Texas who else is from somewhere far away we have any California yeah awesome thank you for coming did anyone come internationally here today travel overseas no hopefully they're online okay all right yeah well we're so grateful that you all are here we are so so grateful and we hope there's lots of time to answer to ask questions and to get answer uh answers to your questions so definitely keep those coming all right yes [Music] should all right um hi everyone uh so my name is Mich I have a
s as as you've heard I joined this department last year um so this is my first year in Michigan and so even though I'm not going to talk necessarily about uh why I chose Michigan and you're going to see some of the things that motivate my research church only today I would want to Second everything that Brar said um I think Michigan is a great place to be uh and I made the choice to come here uh because of the things that I saw in Michigan so I urge you to consider where you're going also in ter
ms of your quality of life um and I think that Michigan and and Arbor and University here and the department all have uh amazing uh opportunities for everybody uh so yeah today uh I am going to talk about some of the things that motivate my research and my work um and some of the things that you would see if you took a course with me which is currently ongoing uh some of the students I think here uh are aware of that uh and scared of me so uh so yeah I I mostly work on spal statistics uh and mod
els with gaussian processes uh but the applications that I want to work on are very diverse and so here I'm going to talk about about how we can use potentially the same uh methods to understand the Dynamics of cancer in the micro environment of the tumor um you know the cell the tissues uh but also when you want to diagnose the effects of climate change in the macro environment right so the the Earth and and these two things even though they are very different from each other actually they are
pretty similar and I'm going to try to convince you and motivate you to possibly uh do work on this and you know there's multiple faculty here members that are also working on gaan processes and scalable gaussian processes for very nice applications um in in medicine so I urge you to also consider that there are many other things that you can do with the same uh kind of structure see if I can yeah there you go um so yeah in the past uh just as to make you understand who who's talking to you I do
n't necessarily have only a background in statistics or mathematics I come from economics and so I'm also interested in the social aspects but currently I'm working on those things that I was mentioning and in my research most recently I've been working on basion statistical methods applications algorithms and software so another thing that uh I do is develop software packages that you can use in our uh open source and some of the research questions that I have is for example what are the local
impacts of climate change we've been studying it uh at a global level is there climate change uh what is attribute what can it be attributed to but I wanted to understand more about so how does it affect uh people and Society locally for example um how are ecosystems changing and some of them may change in different ways and so maybe if you take the average you see no change but then if you look locally you see many changes um same thing in terms of what is the impact of poor air quality on heal
th um last year there you know actually every year in parts of the US there's fires and so people breathing a lot of poor Air U from smoke from those fires how is that impacting them um and also you know other topics such as the interaction effects on health outcomes and the other thing that I was mentioning how does cancer uh affect the spatial relations between cells right so um one thing that I just heard was the importance of AI methods but what AI methods in medicine cannot do yet I mean th
ey can make predictions and frequently they have very good predictions but can they explain why they make those predictions and so one of the things that I want to do in my research is being able to develop mod models that can explain can tell you can have your as a human learn uh why things are happening right because that's how we move on to the next step in research like to get new ideas new directions so first let me start with uh some some um climate change Diagnostics right so this is one
part of the thing and mostly what I've been working on in my past uh recent past uh so the questions as I was mentioning is how do ecosystem develop in terms of climate change uh what is the effect of human intervention and so on and so in these situations we have multiple like a multitude of data sources right so we have weather stations you know on the ground we have satellite imaging that take you take pictures at regular intervals of Earth we have remote sensing such as lighter uh sensors th
at take like depth measurements camera traps to see whether there's animals around and citizen scientists meaning just people who walk around and take notes right so many many uh data sources and one example of this uh uh this kind of set of problems is you know again air quality monitoring and so we have a weather station that measures also the quality of air we have fires and this is going to have impact on human health and so this is one uh topic that I am interested in but also so relates to
the research that I've been doing and the models that I develop um and see here's an example of the kind of data that you would be able to see and that my students in in in 696 spatial statistics do see all the time and I ask them to work on on developing methods and fitting models for these kinds of data sets you know there's multiple sensors here in the Pennsylvania area and we have like air quality levels um and then you can use these to assess how people are breathing in this air and having
uh issues or non-issues for their lungs example and the other topic that I'm talking about here uh is the like trying to understand the cancer micro environment and so the thing that I'm saying is basically that even in this case we have multiple sensors right you can imagine that make I'm making a parallel with what we saw before so here we have U Medical Imaging such as MRI or CT Imaging we have tissue tissue biopsies in which take a little bit piece of of you know humans tissues and then tak
e pictures from a microscope and we have Laboratory Testing in which you know we have other sources of data and once again even in this case there's a spatial aspect as you can see here um this is a completely made up a image of two tissues um this is AI generated on the left we have uh what AI imagines as being healthy tissue and on the right some non-healthy tissue and you see that the even AI knows that the unhealthy tissue is um different you know from a spatial perspective relative to healt
hy tissue um even in this case if you look at real data so this is now um on the left you see like the uh markers of different cell types and you see that there's a special um structure in all of these uh cell types and how they are patterned in in in in in the tissue and so the idea even here just like we saw in the air quality monitoring case is that we can use special methods to model these um data and find uh explanations for how things change and how things are and why they are that the way
that we see them and so how do we link them so the similarity as are of course that you know we have in both cases data with spatial and temporal coordinates uh the data are multivariate so it means that at every coordinate we have multiple uh sources of information such as you know multiple cell types maybe or multiple uh environmental variables and in both cases is we have high resolution sensors we have massive data sets and this is going to be a problem because many of the methods we develo
p are actually not very scalable meaning that they don't they don't extend to very large data sets and so one topic of research that I've been interested in is in how you develop methods that do scale to massive data sets and that's important and very influential uh and we have coverity information you can imagine like the subject information in case of Medical Imaging or the environmental information in case of the environment and we have complex interactions of course and multimodality as I wa
s I was saying you know that just means that you have ground level data and then satellite imaging and they tell you kind of the same information but in a different way but the differences are that we have just one planet Earth so that is the one subject that we have in the case of the environmental uh studies whereas in the cancer studies we have multiple subjects um and then in in the environmental case we have observational data whereas in in in medical studies we possibly may have randomized
trials in so treatment effect that we want to measure for example and the treatment effect may be different according you know in space that's uh not very simple to estimate and so in all of this what I'm saying is you can use basion methods uh andas methods are the glue that can help you um Bridge like these areas that look different but actually I'm I'm arguing they're very similar and in in aasia method I'm not sure how many of you are how many of you are familiar or have heard about measure
methods many of them so yeah that's very good so we have a probability model for the data uh that's like our instruments right so the the the the the the okay English a second language I don't remember how things are called speaking of Internationals um yes where you put the colors and the instrument you used to draw okay that uh but it's not the there yet right so we have PR uncertainty palette the palette yes the color palette yes of course uh and the the the what's it called the brush brush
thank you it was not that difficult right I I think I think I should I should have known uh but see I am I'm I work on other things not on not an artist I mean yeah I let I AI do everything here so um so again we have prior uncertainty so kind of an idea of how things should look like before we see the data right and then we observe the data we see which color we are supposed to use and then we draw like you use the uh brush to draw whatever the um solution is to the problem right so the posteri
or is just the uh upd updated uncertainty that we have about model parameters that we don't know anything about except for the prior information so uh and you know here I'm not going to uh go over it too much but essentially we may have a um data for subject at some location so you can imagine that this is potentially the pixel of an image or for for a subject and if we only have Earth then I is just Earth and then we have a model for all subjects we want to explain the data with some parameter
that we don't know we want to estimate that and then that parameter can be simple such as a number but also a whole function such as a surface right so that is uh when for example we want to estimate the effects uh in space that effect in space is you know a big image uh and so the prior and surfaces that we use is typically a spatial stochastic process uh such as a gaan process uh and so then you know in our basan um framework we observe the data so here would be the data for for example multip
le subjects um I think in the medical setting and then we uh update the uncertainty that we used to have the prior using the data and find the posterior and you see here basically I'm applying the same method right so we have a rough idea of how the shapes should be and you know the but we don't know how to color them and then we put color on that that's so that's a representation of how a how a basian uh framework would work and with once we get the posterior so the goal of a basion model is to
get the posterior once we had the posterior we can do everything we can do test of hypothesis we can give you uncertainty intervals we can make predictions everything we need to do is inside the post so the goal is always to find the posterior and in the um satellite imaging case so for example in this case you have a situation in which you have snow and uh trees and you know that the water cycle uh especially with like a CL in a situation of climate change is very important for not only um you
know directly to understand what the effect of climate change is but also Downstream the effect on people you can imagine in California this year they had a lot of water but the year prior they we're having no water at all and then if there's a lot of snow accum accumulation then there's going to be water in the spring if not there's n and so that impacts not just the local uh you know areas that have or do not have snow but you know overall society and so in understanding those changing and ma
king predictions is very important in the future and here we use scale of bibas Gan process regression methods to basically um be able to analyze these massive sources massive quantities of data in this very complicated setting in which there's a mul multivariate outcome and you get all of the things that I was mentioning including uncertainty intervals and and so on and in the uh this is ongoing research now this is in the tissue biopsy case we have multiple subjects and with some people were a
sking about collaborations with uh external departments so this is a collaboration that's ongoing with the center for uh cancer bioinformatics uh at the med school I don't remember its name ccmb um so there's a PhD student and a post loock there that are working with me to develop this uh software package that is basically an extension of the software that I was developing earlier for um the environmental Sciences uh to this situation in which you have multiple subjects and you want to understan
d how the two more micro environment changes uh in time and space um for the different stages of cancer for example right so that would explain how cancer impacts cells and the relationships that they have with each other and so this is like initial uh steps in fitting those models and trying to understand how things work um you know again uh this is ongoing research we are planning to have something out by the summer um but yeah so if you were coming to Michigan you would expect you know not ju
st with me but of course with everybody uh that you can be exposed to a variety of problems and a variety of people to work with and so I think that's very appealing I think that's very um advantageous for your career in the future because you can always say that you have had experience in a variety of things as opposed to just working on a narrow set of things so again why aan Paradigm this is like trying to convince you to to use this not necessarily to convince you to come to Michigan but a l
ot of people in Michigan bans no so it's a flexible strategy for complex data it's based on the math of probability uh it forces you and my students know that it forces you I force them to make clear assumption uh on you know what your prior is what your model is and that's always the question that I ask students and some of them can can attest to that that they need to be very accurate and specific and detailed in how they Define everything because that's the nature of a basion model to be like
very clear um and again this forces the user you know of a software package for example to think about what you're doing the science of what you're doing it's not just about running a couple lines of code and getting the answer you kind of need to understand what's going on because because it's you know complex but also flexible and it's very interpretable so you should be taking advantage of that and that leads to reproducible science I mean there's always talk about having this crisis of repr
oducibility that you know people run studies and then others run the same study you know trying to figure out exactly the same steps and then the results don't reproduce and so this being very transparent very clear that it's the nature of aasan u a basion model allows you to be hopefully more reproducible and so you know more impactful in the longer term future and so if you want to have like some uh references about what I've been doing um here's some um then I think I'm done with the presenta
tion today I hope I have motivated you enough uh but you know you can talk to me or anybody else um here I think it's going to be a very good move if you move thank [Applause] you do have questions I hope I haven't scared you with uh with my talk about you know courses yes please oh we have another hi I'm Jennifer so you explained earlier that you're doing stuff both with like the cancer stuff and the climate change and then you said for the cancer stuff you were collaborating you know like like
in the health departments and then for the climate change who are you collaborating with for for the those days thank you for the question so yes um as I've moved the year uh basically my initial step was to try to establish some uh relationship ship with people at Michigan right because it's important to to have like uh your foot in like multiple places right and so and so the thing that the first thing that I uh that I could find was actually this um collaboration with people at uh the med sc
hool um I think there at Michigan medicine computational medicine and bi computational medicine and bioinformatics um and so because they have access to data right so for me the goal is to have access to import to cool data that inspires new models uh and so that's what I did in terms of the climate change part that's that's that's all of the collaborations that I had before coming to Michigan so I have my post advisers are uh part of that I have a collaboration that was ongoing with the uh Geor
ge Washington University um that was like related to modeling the um I have it here yes uh the the um change es in the speed at which trees become green during the year as a as a consequence of climate change uh and and so that's another part um I hope that answers your question hi my name it's Emily Maron um I have a question you mentioned earlier something about um it was particularly large language models right that you were trying to get them to be more transparent so I mean I'm yes ask the
question please so um I don't want to interrupt you how exactly do you propose to do that cuz I I do a bit with large language models and some other models but um they're very very not transparent at all about how they're thinking exactly so I'm wondering how you're proposing to make them uh expose how they're coming to their conclusions aside from probability Maps I completely agree with you that they are very obscure in how they give you the perfect answers sometimes right so there's I think t
here's a paper that was published recently in nature uh I want to say uh I think it was a group of Stanford researchers that used a you know a very large data set of images and they were able to predict uh or make predictions about cancer development using images right and so I I read that and I'm like Okay cool so 99% accuracy in making a prediction on your cancer stage based on images that's very accurate but you know how did you do it right uh and so I don't I don't think that it's necessaril
y easy to like reconstruct an AI model so that it tells you exactly how it's doing those things because it's the nature of those models to be deep you know and every layer in those uh hierarchies means that you're losing interpretability because there's like a billion parameters that all of them have like different uh effect effects so what I'm what I'm actually saying is that we could potentially use the results of you know those predictions consider them as our data and make them uh look like
noisy data and then try to you know because now we have a lot more uh you know the data can be we don't have to just rely on the human labeling of images and cancer stages we can use a large language model we can have a massive data set of images and then we can fit a basion model to those Imes you know taking into account the nois what did you say at the beginning analyzing the output and compar it to the input and what a human would do and then you're kind of like interpolating so the the AI m
ethod the ai ai model is typically trying to do that right so you give it a training set and then it's learning from whatever the human does and it's going to do try to to to mimic the human as much as possible uh but the human you could ask and the AI model you you don't right and so the AI model will generate generate a ton of data though whereas the human would would take a long time to generate the same amount of data so now that we have a lot of data we know the answers to this question it'
s pretty accurate we can trust the AI model that's going to give us something that's maybe noisy maybe not always 100% accurate but pretty close uh so that's that's nice and then we can just use that data and reanalyze that data and kind of try to give an explanation of why the the AI model came up with those answers and I think ban methods are actually you know basan methods and I mean um interpretable and flexible basan methods that you can interpret and you can understand understandable measu
r method there's a whole line of research on that um can actually do that for you right so they can tell you here's why the AI model gave you that answer I think that's that's a very potentially powerful thing that we can do because people are expecting that as a you know in the future I think it's not just going to be limited to oh yeah sure here's the and you can see that when you use CH GPT right so sometimes it's like okayy thank you for all of that U you know cloud of text but I mean it sou
nds okay but maybe I want to have like some input more which is why we need prompt engineering then so but y I hope that answered the question all right thank you so much Mel that was excellent Graphics just phenomenal as usual that's AI thank you so much prompt engineering so um we are running a little bit behind schedule in usual admitted student day uh fashion so we're going to take a quick break now to make sure that you get to use the bathro stretch your legs and come back so you can get fu
ll attention for our student research presentation um and then we'll move on from there so if um you'd like we'll come back at 10:35 um which is just a little under 10 minutes there is still food in this room if you're still hungry or need a snack otherwise if you go out here turn left and left again down the hallway are bathrooms um so we'll see you back at 10:35 thank you e e e e e e e e e e e e e e e e e e we don't get too far off schedule um there will be plenty of time for chatting and and
learning about your neighbor so um I don't want to I don't want you to stop that I do want you to stop now but I don't want you to stop that generally okay so um our next presentation is from one of our uh senior PhD students um so we have the pleasure to listen to Jeff okamoto right now and then we'll um go right into the admissions part so um thanks Jeff go ahead everyone I'm Jeff uh I'm a PhD student here um and I'm going to be talking about um some of my recent work on uh integrating transcr
ipton wide Association studies and calization Analysis so a quick show of hands anyone ever heard of either of those before one okay that's more than I thought okay so uh I'm going to spend the first few minutes uh filling you guys in on on what exactly colocalization and transcripton white Association studies are uh then I'll talk about my uh my method and talk about a um a quick real data application so the broad research question that we're interested in here is how can we study the genetic m
echanisms of of complex traits uh you can think of a complex trait as something like your standing height or maybe like a disease status uh so one of the most popular ways that researchers have done this recently is called a genomewide Association study or Goos um so these allow us to see basically the snip snip level underpinnings of a complex trait uh so what's a snip basically it's just a single substitution of a nucleotide at some place in your genome and the idea of a gwas is basically to p
erform a simple uh linear regression of your complex trait levels um on your Min minor alal count uh at that snip uh and if you do that basically hundreds of thousands to millions of times across your genome you end up with a plot like this um where you have basically an association for every snip in your genome and you have the uh the association strength here on the y- axis and the location uh on the x-axis and we call this a Manhattan plot so the next step is basically linking these genetic a
ssociations to the genes that they influence which we call putative causal genes here um and this is actually pretty challenging um so traditionally researchers have relied on our prior biological knowledge of uh genes in proximity to those Goos Losi uh so we have the same Manhattan plot here um as the last slide and I've revealed to you that it's a Jos for serum testosterone levels in females um and you can now see that uh they've annotated um the top G based on that previous biological knowled
ge so uh fairly recently a new class of methods has begun to emerge called uh mechanism aware PCG implication methods and the idea of these methods is to take what we call multi-omic data uh and integrate that uh to identify PCGS and uh reveal their underlying mechanisms so some examples of common data types multiomic data types that we uh often integrate are transcript omic or gene expression um and proteomics uh and some popular methods that we use to do this are called calization analysis and
transcrip Association studies or TS so what's a calization analysis um so the goal here is going to be to identify or to sorry to determine uh whether genetic variants that are causal for a molecular phenotype so gene expression for example and we call these variants qtls uh overlap with those that are causal for your complex trait so your gwall crate so here I have a little cartoon uh we have snip X which is causal for both your gene expression of some Gene Gene a uh and your complex trait so
this represents a calization so this is actually a fairly complicated task in practice and I'll show you why with these two cases um so case one on the left we have our true co calization meanwhile case two on the right we have separate uh qtls and jets so we don't have a calization so it turns out that if snip X here is uh perfectly correlated or in genetic slinga we say in perfect linkage to equilibrium or LD with snip y then these two cases are going to be indistinguishable so in response to
this problem uh researchers have developed basian methods to try to resolve this case um if snip X is in weak or moderate LD with snip Y and basically the idea is that you can quantify the uncertainty in the presence of calization through what's called a gene level colocalization probability so that was one way to link genes to your gwos trait now I'm going to talk about another way another way which is called toas so toas is a form of what's called instrumental variables or IV analysis um which
uh is designed to use observational data to test for causal relationships from some exposure in this case gene expression to some outcome in this case your complex trait so we're interested in whether or not this red on this graph exists um so this uh this form of analysis comes actually with some pretty strong assumptions because we're using observational data um one of them is randomization of your genetic variance one is the relevance of your uh your instruments or your genetic variance to y
our uh molecular trait so just your qtl strength and then uh the final one uh which is often the most difficult to validate uh is the exclusion restriction just being that your Snips uh can't be causal for your complex through any pathway other than um the exposure that you're considering in this case or gene expression so in practice the way that you do a TS is you have some expression reference panel um on the right here that you use to train a prediction model um then you use those prediction
weights in a separate gws data set to predict your expression across the transcriptome you end up with your transcript and wide set of predicted expression then you correlate all of those uh all of those predict predicted expressions with your GS trait so you end up usually with like a transcripton wide set of Association statistics like a zcore um so uh just to recap a little bit here calization analysis was a way to probabilistically quantify or overlap between causal qtls and gwos hits often
returning a calization probability for each gene and Tas was another way to link our genes to our trait um this way tests an association between your predicted expression and your complex trait uh and you often get a zcore as output for each gene um so uh recent work has actually found that these two types of analyses can be complimentary when applied to the same data um in particular they don't always implicate the same set of genes as causal um and there are certain uh underlying biological f
actors that can actually explain this um so uh for for example if you have a strong colocalization but weak t-w signal this can indicate the presence of what's called horizontal Pat tropy so we have here on the left what horizontal P tropy looks like where you don't actually have that Arrow from your gene expression your complex trait so this is not a causal Gene uh meanwhile uh we're looking for cases like on the right where we have vertical P tropy where you do have that effect um and on the o
ther hand if you have a strong t- signal but weak colocalization this can indicate what's called an LD hitchhiking effect effect so in this example you have snipex your qtl that you're using that's uh an LD with uh some direct effect direct effect snip snip y um and basically this is inducing a false or spous TS Association um and you can see in this case you don't actually have colocalization um so going off of that uh this previous work offered a strategy to try to reduce these spous TS result
s um and the idea here was going to be to basically filter out uh basically filter your your t-s results using a calization probability threshold so the idea here graphically is going to be to superimpose calization evidence and hopefully that'll get rid of cases of of SP where basically you don't have a true causal effect from your uh Gene to your trait uh and hopefully you'll just be left with uh your true scenarios where there there is an effect um so the major drawback of this approach uh is
it's it's pretty ad hoc and you you lose your uh your your uncertainty quantification in your your uh PCG implications okay so now we're going to get into our methods overview uh and our more focused research question now is going to be how can we take our evidence from these two types of analyses t-w and Co localization uh and probabilistically integrate them to implicate our PCGS so we're going to start with a model um to motivate ourselves so the top equation is for our gene expression of ou
r Target Gene our candidate Gene um and this is going to be a function of the genotype Matrix G and our true eqtl effect Vector beta e um our bottom equation Y is for our complex trait or GW trait um and this is a function of our uh true Gene to trade effect which which is what we're interested in uh gamma um as well as this extra G beta y term which represents the uh plyopic which are not mediated by our gene expression so previous methods have actually also considered a similar equation um but
these impose additional assumptions to try to identify that uh gamma term um for example they assume that one method assumes that the the two beta vectors beta e and beta y are uncorrelated uh and another assumes that your ply Tropic effects are constant across variant so in practice these actually aren't very reasonable assumptions um they're often violated in real data um so to try to um get around making these assumptions we're going to focus more on testing whether gamma equals z rather tha
n estimating it um and as sort of a launching point we're going to note based on this equation that our causal eqtls are going to have to be colocalized with gwos hits based on this model and basically to to show you that just sub in the the E equation into the Y equation and you can see that genetic variance with nonzero beta e are also going to have a gws effect so our method which we're going to call integration of T and Co colocalization are intact basically is going to incorporate that obse
rvation um into an empirical Baye framework to implicate PCGS so a quick overview here we're going to first form a base factor from RT y zcore to represent uh marginal likelihood then we're going to form an empirical B is prior uh using our colocalization evidence so that's going to take the form Pi F of p p Colo where P Colo is our colocalization probability we're going to uh estimate pi as a unconstrained T prior uh and the function f is going to satisfy two properties the first being that it'
s monotonically increasing with our calization probability and the second being that it's thresholded at some value T just to make sure that if we have a very strong spus t signal we don't want that to overwhelm a very small calization probability uh and then at the end of the day we can use base rule to form our posterior okay so summing up here again um our goal was to estimate a posterior probability of causality essentially for each gene um we used an empirical base framework to do this that
involved uh converting a t zcore or a base factor and then we also formed an empirical base prior based on a calization ation evidence um uh okay so now I don't have time to show you how well our method Works in simulation um but uh I'll jump into a quick real data application um so this for this application we're going to look at four different gwos traits so we have serum urate igf1 uh and testosterone separately separately for males and females um and the reason why we chose these traits is
because we actually have a lot of biological knowledge ahead of time about these traits so we can actually use these to see how well our method is working um so these traits can be further broken down into subpathways which is in the second column to the left uh and each pathway is annotated with a number of core genes or genes that we just know are relevant to these these uh complex traits from our prior biological knowledge so that's annotated to the right of the pathway name and then the midd
le column is going to be the number of genes that we have implicated uh based on the proximity plus knowledge approach so proximity plus knowledge is just basically the approach that I talked about in one of the first slides where uh basically you implicate any Gene that has a Goos hit um basically in proximity to its coding region um in your genome so to the right of that we have the number of genes implicated by intact and we're integrating uh multi-tissue gene expression data for this uh for
this implementation um and then then finally on the right we have um the overlap of the cor genes implicated by both approaches um so what I really want to point out here is two things uh first of all we have substantial overlap between the number of uh between the genes uh implicated by each approach you can see that on the right uh and then also uh there are several cases where intact implicates core genes that we miss using the standard proximity plus knowledge approach uh so we're going to c
onclude here that uh intact uh is a complimentary approach to the standard okay so just uh wrapping up here so uh intact um was is basically a new method to explicitly link genes to complex traits using uh the transcriptome uh the the main idea here was to protect against LD hitchhiking Effects by constraining our T results using calization data um and then we just saw that we were able to complement the proximity Plus approach um and some uh extensions that we've been working on recently are ad
apting intact to consider uh additional molecular phenotypes so protein levels for example in in addition to our expression um and then also to focus more on uh trying to estimate those effects rather than just uh testing for them um so I'd like to acknowledge uh my advisor um as well as our co-authors on this work um and if you're interested I encourage you to check out our paper and our our our pack package uh thanks for [Applause] listening all right um that was awesome thank you so much Jeff
if you have questions for Jeff I'm going to ask that you find him um after our uh admissions piece so we can get moving so you can get to the faculty but at that time hopefully Jeff will be around and you can ask him questions or you can email him um so thank you so much so um welcome I am the PHD committee uh admissions committee chair my name is Kelly Kidwell I'm also a professor here in the department um and Associate chair for academic Affairs and I'm just so grateful that you all came here
so thank you so much for applying to us for considering our program we're so glad that you made the trip here um and we hope that we can convince you that this is the right place for you a little bit more about our department um we have a very large Department both in student and in faculty but this is really an advantage this is an advantage for you it's awesome for us you don't need to know what you want to study you come here you start taking classes you meet everyone and you'll figure it ou
t because whatever you want to do we've got somebody here that does that and they're an expert most likely a national International known expert in what they're doing um and so that's really one of the big big advantages to a larger program also all the students here you get a really great pick of friends right there's a lot of options so it's a really great thing to have a nice big program and a lot of research money that comes in to support a lot of awesome projects so we had a record-breaking
year yet again I've said this year after year but somehow we keep breaking records of the number of applications so 981 applications to our M's and PHD programs um split about 2/3 1/3 across Masters the two programs and our PhD um so we've been busy we've been reading all about you and we're so grateful that you found our program um we have made 485 Master's uh acceptances now we do not expect that many students FYI okay don't get scared we're not increasing our program uh fourfold uh we antici
pate numbers between 60 to 80 in our uh biostat Masters and 20 to 30 in our HDs Masters okay and then we made 51 fully funded PhD offers and we anticipate somewhere between uh probably 20 to to 30 or so students um so really nice cohort sizes um really excellent excellent program and and a lot of students that you'll be joining that just provide so much great information for you um a little bit of background in terms of our last cohort what sort of background they're coming from and probably mos
t of you um so we definitely have a lot of students coming from our to our masters with a math or Stat or bio stat background that's the largest uh representation of backgrounds in our program but we love all types of backgrounds right diversity in terms of um of who you are and what you've studied really helps the the um program and the science and just gives it so much more uh so much more to share so we still have we have biochem backgrounds engineering computer science economics business and
some other backgrounds um represented so as long as you had those three prxs right we're really excited about getting you here our phds um primarily have that math stat background we do have a lot of PhD students from our own biostat masters um and then coming from other Master's programs or directly from undergrad usually uh with a math stat background or data science some Finance obviously more a little more on the quantitative side with our PhD program um we have our two Master's programs no
w so we have a MERS of Science in biostatistics and we have a MERS of Science in health data science um both of these excellent excellent programs you had to choose one or you applied to both and you got admitted to one or both of those um you know a little bit about the difference so bio statistics is like the traditional biostatistics um whereas HDs is going to have a little bit more focus on computation um we're trying to develop biostatisticians whereas Health Data science is maybe more Heal
th data scientist however those terms are somewhat synonymous these days so I think you could definitely call yourself a data scientist coming out of our biostat Ms um both of these programs there's high high overlap um a strong foundation in statistical Theory they share most of the core courses um a lot of the elective courses so if you're in the traditional biostat you can take elective courses from the HDs program if you're an HDs you can take traditional biostat courses so really great um y
ou know combo uh lots of shared shared courses here um either of these Masters could be terminal degrees set you out into the world with an amazing career um alternatively if you want to have this as a stepping stone to a PhD here or elsewhere uh these are great great degrees for that so a lot of overlap a little bit just more emphasis on required classes being based on computation and big data and HDs whereas those aren't going to be required but they could be electives in biostatistics Ms both
programs are 48 credit hours and they are residential here over four seme semesters so we don't have classes in the summer uh but you will be here in the fall and what we call the winter some people call Spring um it has been quite springy lately uh but it's going to be your two primary semesters and two years three qus of those credits are going to be um really bio statistics or health data science classes um and the other are going to be filled with electives that can come from other discipli
nes we do require 3 hours of an epidemiology course and a one credit or one hour of a public health course to give you that nice rounding of of uh Public Health given we're in the school public health um but otherwise you can fill in your electives with courses um from biostat from statistics Department in lsna from computer science information um Sciences engineering a lot of our students take those courses so once you're here at Michigan you can take all the other Michigan classes um we're rea
lly hoping you know to get you to be a great team member um and and potentially a a really well you know give you all those skills so that if you want to leave with your Masters you will be ready to jump into that professional world the biostatistics for the MS coursework includes these core courses so these are required for all individuals going through our MS is a series of probability Theory and statistical inference biostat 601 602 uh you'll take that your first year fall and winter semester
ERS um and alongside of that you'll take this uh statistical methods are more the applied regression courses biostat 650 651 um along with those in the fall and winter semester we also have bioet 653 which you'll take your second year um and then it all culminates with this Capstone course so we don't have a master's thesis uh or a master's exam what we have is this Capstone course biostat 699 that you take in the second semester of your second year that brings everything all together um and so
everything you've learned that you might be like why did I learn this Theory What What In bioet 6.99 they're going to say hey here's a project figure it out and you're going to say oh I need that thing oh here's where it comes together oh and now I see how this how this applies to this thing right so it's all going to come together and really prepare you for that next step um and then we have a a bunch of biostat electives so for example like survival analysis or clinical trials um or non-param
etric statistics categorical data your your choice on those um and then you can choose some additional electives that could be outside of biostatistics the HDs coursework is similarly going to have that probability Theory and statistical methods those four core courses 601 602 650 651 um but instead of 653 uh we require these um uh Computing and machine learn learning courses so 620 625 and 626 so Health Data science Computing and machine learning um and then we instead of$ 699 we have sort of a
health data science equivalent which is biostat 629 and so that's going to be very similar bring it all together you're going to work on case studies and apply all those those methods that you've just learned um we have a required biostat elective from a specific list for this um degree and one specific Computing elective that you must fulfill from a specific list um for this degree and then the additional electives you can choose quite broadly our PhD program is typically about somewhere betwe
en uh three to five to six years depending on if you have your Masters either from our program or a different program or if you're coming right from undergrad it's probably going to take more on the five to six year span um so if you're coming right from undergrad or even from another Ms usually you take our master's coursework uh to begin so you take that same 601 602 650 651 all those courses you do two years of that Master's coursework uh to as a foundation um and then once you've done that y
ou could take our qualifying exam currently that looks like one exam that's based on those four core courses um or maybe $6.99 as well so 601 602 650 651 um and then you have about a year left of coursework for the PHD program um which includes you have to take an advanced calculus class or real analysis class um along with a higher level statistical inference set of classes 801 802 um we require stochastic processes um and then you can choose additional biostat electives once you've passed the
qualifying exam you've completed your coursework you're now a PhD candidate um and this is really exciting this is where you get to go around to our faculty and say hey are you cool enough to be my mentor right you get to interview US you get to see do you want to work with us do you want to study with us um can you can you like hang out with us enough for the next few years to do your dissertation um that's that time that you decide that and you start working on that dissertation and that could
take anywhere between two to four years um we're really hoping that from our PhD program we're building up your skills your foundation your statistical Foundation um to be a real great leader um we have this new coursework that we introduced this past year so two years ago we started as a pilot this last year it's it's uh full-fledged so all of our first year students we really suggest highly highly suggest that you join our classwork bioet 611 in the fall and 612 in the winter um this is a gra
d school and professional success skills class so this is really going to complement all of the statistics and the foundation that you're learning in the classroom with all those other skills um that how do you be Su how are you successful in the the classroom how are you going to be successful on the job market how can you communicate well um how can you deal with all the pressures that you have uh this class is really just um an amazing opportunity that we have here it's one credit there's no
homework um it's just to come one hour a week and to listen to various uh professionals across a variety of fields uh to help you Excel as a whole person not just as a biostatistics student um so it's going to help you survive and thrive as a biostat student um but then also help prepare you for internships in the summer or job opportunities whether you're a master's or a PhD student um this is an excellent excellent class so highly highly recommended um a little bit about our funding sources so
all PhD students are funded several Master students were um offered funding although not that many as you saw on that initial slide um and if you're offered funding what that means is that you your tuition is fully paid for and you receive a stipended and in return you do some work so you either get a graduate student research assistant ship which you can see is the most of our students who are funded um and so you generally work as sort of like an apprentice with a faculty member um or you mig
ht be a graduate student instructor or a teaching assistant um there are less of those because we do not have an undergrad program so you'd be helping with the um foundational courses for the other school public health disciplines for learning biostatistics or some of our uh B some of our earlier Master's level courses um we also have two training grants so we have the genome science um training program and the cancer biostatistics training program and so uh several of you might have heard from
um some of the faculty who lead those programs uh for interest in in those uh training grants in addition we have a number of tuition Awards so some of you might have been offered tuition as opposed to full funding uh we gave out a quarter I think actually this year we gave out 50% and full tuition to some uh individuals and so that means you don't have to do any work for us you don't get an additional sttip but your tuition or part of your tuition will be paid for by the department um so if you
weren't offered full funding um you can look at this uh great guide to student funding from Michigan website we included that in your letter um and so you can potentially pursue other uh GSR or gsis in other departments if they're listed on there you're welcome to pursue that at your um you know on your own um in addition we have something called statcom which isn't methods research but it is collaborative applied um uh opportunity that you can get involved in seeing how you use statistics in t
he real world um to help nonprofits and so it's a really great opportunity it's a volunteer opportunity and you'll hear more about that from our students um we have a course called biostat 610 and this is where you can talk to a faculty member and see if they're willing to do this with you it's called readings in BIO statistics um and you can just essentially have like a weekly meeting with that faculty member start reading in some area maybe working on a project with them um and so that's not a
paid position however you get some opportunities to do to to start to figure out what research might be like with someone um several faculty members will hire what we call temp hourly workers um and so in that case you're not a GS however you could start to work on um most likely a collaborative project or or something that a a a faculty member has for somewhere usually around 10 hours up to 10 hours a week um and so you can have some opportunity to work uh with someone in that way and then the
summer time is a time when our faculty will more likely hire graduate student research assistant ships uh you're a little bit cheaper in the summer um and we have a little bit more time to breathe and help with uh these projects um and so once the kind of like winter spring comes around you can check out if any faculty have these opportunities um and apply for those also um obviously there are many many internships outside of University of Michigan for the summer that we highly um uh you know e
mphasize to our students to to apply for those and we give lots of resources for the for applying for those getting ready for those particularly in that biostat 611 and our amazing Career Services office in the school public health our students from the master's from the PHD have just amazing job opportunities so our departmental reputation the importance of biostatistics and health data science uh the strength of our students and the strength of our alumni Network um which is vast um all across
the world are really really helpful in getting our graduates into amazing positions um so our students go into universities academic research centers they also go into government they go into industry they go into consultation jobs uh technology jobs across all different sectors where someone might be interested in someone with quantitative skills um we post notices of positions when we hear about them we send them around um we also host recruiters so we had a uh job fair uh just recently uh at
the beginning of the semester to help um we also have something called the alumni Spotlight where we invite alumni back to give a talk about how they got to where they are what they're doing and then they make uh appointments with students to chat with them uh biostat 611 as I've already harped about really great opportunities um given there and then our Career Services office is wonderful so here's just a little quick image of where some of our students or graduates have gone uh Brar said a fe
w of those uh amazing female students where they have gone but you can see over here we have we have academic centers we have research centers we have technology uh we have pharmaceutical companies we have software companies right we've got it all represented really amazing places you know great jobs um all at the fingertips of our graduates um and so you can see this is a little bit older data from 2015 to 2020 but about a third of our students from the master's program going into industry um n
ot quite uh half of them are you know a little like about 40% going to PHD programs um almost 20% going into research universities or Hospital positions and uh 4% going into government or nonprofit positions our PhD um almost 60% going into industry and Industry here we're kind of lumping together you know like pharmaceutical technology um places like that uh about not quite a third um from 2015 to 2020 going into Academia 10% going into government and nonprofit positions so um just really excel
lent opportunities really great spread wide variety of opportunities where our graduates are okay so the deadlines you guys have not quite a month to make your final decision but if you know earlier please let us know as soon as you know we would love to welcome you and congratulate you on on your decision we hope that it's here but regardless of where it is we want to congratulate you and and be really excited for you making that next step in biostat or statistics or whatever that step is for y
ou um once you have made that decision it's official um right after we have that kind of information from everybody we're going to send out if a position questionnaire so if you've been accepted with funding we're going to send out a survey to ask what are you interested in for your funding are you interested in teaching are you interested in research who are you interested in in who's research and what type of research um and so we use that information and we do a a departmental um matching sys
tem so that we take your information we take the faculty with funding and we do our best to match you and the faculty with one of their top choices okay so um that will happen and after you've accepted after we know the numbers um and the funding information um so the late spring and summer you'll find out who you can work with if you're fully funded and then we'll have our new student orientation so the week before school starts we like to bring you in and do all this over again remind you um a
nd make sure that you get to meet all of your fellow students and have great food and fun our first day of classes in the fall is August 29th um so if you are a master student and you weren't accepted with funding which is the majority of Master students um you are potentially considered for departmental funding if opportunities arise so um if we have extra uh teaching assistantships or research assistantships um or tuition money um which you know sometimes happen that there are the few addition
al positions um then we will look at your information and we can potentially offer you one of these things um also as I said there are these temp hourly positions um however I don't want to get your hopes up too much so there are potential opportunities however they are few so if you are an unfunded Master student we do want you to prepare for paying for that Masters for um all four semesters so I just want to leave with our shared Public Health values we believe in compassion we believe in Inno
vation we believe in inclusion and we pursue impact in the School of Public Health and in the department of Bio statistics we are just really really excited and Incredibly um really just look forward to to your um hopeful acceptance into our program if you have any other questions we'll give a few minutes for questions before our faculty come in to introduce themselves otherwise you can ask faculty I have a room so you're welcome to come ask me um or you can send emails FMA for Master's question
s Nicole for PhD questions or I'm happy to help figure out where that question should go so with that are there any questions for me yeah um let me give you this starts a okay maybe August 29th was this last year so you should definitely look at the registar for that correct date thank you any other questions about yeah are OPP for those admitted to the PHD program coming out of just a bachelor is there any opportunity to like test out of certain courses I I'll I'll repeat into the microphone br
iefly um are there any opportunities for incoming PhD students without a masters to test out of some of the Master's courses oh okay yes there is um so uh you are welcome to submit coursework um to our program our curriculum committee will look at that um and decide if you can wave out of any of our courses I will say that the majority of um individuals who come directly from undergraduate usually don't wave out of the courses for the PHD program um it's more likely that if you came with a maste
rs from a different program you might be able to wave but you should try if if you think that you could potentially do that yeah let me get you the the mic first hi I was wondering if PhD students coming in without a masters actually earn the Master's Degree um when they finish all those requirements um great question Nicole do the master do the PhD students without a masters do they technically earn their masters of the same master and you can apply for an embedded Masters which is the same it
comes in as the Master's Degree so you will have an opportunity to apply for that whenever you finish that but it's it's it's a master degree yes great you don't have to do that but you can do that is that what you're saying yes yes um if we got like a tuition paid but we didn't get funding does that mean we don't have to do anything if we got tuition paid that's right in terms of you don't have to do anything in terms of a gsra or a GSI that's right your grades up correct yes yes come and enjoy
go to classes do well yes yes all right here hi uh this is Indra I have a question regarding the uh funding um so when we will know we are considered for the funding uh after we accepted the uh admission or after after April 15th um so your admissions letter would have said if you were accepted with funding um so you would know right now now if you are accepted with funding or not and then you will know the specifics about the funding in the late summer um before the fall after we get that info
rmation from the survey so do we have any eligibility uh uh basis which we'll be eligible for the funding or uh how how the process uh is designed sure so when we're looking at uh applications in uh the admissions process so the PHD is automatic funding so anyone who accepted to our PhD we know we're going to fund you and then in terms of Master's very little students were offered funding um and there's not a specific set of Eligibility criteria but um you know we're we're just looking for um th
ose in which we think would Excel and and which we think would add to our program um and and might be able to come into a gsra or GSI position all right I'm sure many of you still have many more questions and again you can ask these to The Faculty that you meet with or you can come with meet with me um but we've made our faculty wait long enough thank you so much for your patience faculty as usual we're always behind um so thank you for coming we're going to do quick introductions of all of our
faculty members um and then we'll have a time period in which you can go to their rooms so they all have assigned rooms um and every 15 minutes we're going to have a little switch up so figure out who you want to meet and then you'll switch to different rooms um if that room is too full go find a different faculty member and come back later okay we have enough time for that so we'll start over here just one more quick announcement for the people online we will be transitioning to a student panel
for the online audience um when the faculty and students exit the room thank you uh for faculty if you can just introduce your name and then maybe just quickly interest uh hi everyone welcome um my name is Nick Cartman I'm a research assistant professor um a lot of my research interests are focused on Survival analysis predictive modeling um with a focus on health policy and kidney disease hi everybody welcome my name is Kathy spino I'm a research professor and um I work with clinical trials an
d data coordinating centers hi everyone uh I'm Jen Kong I'm a professor of B statistics and uh my research interest is in machine learning Bas method and with application in imaging hi everyone my name is weall and I'm a research assistant professor and my research uh focuses on the mediation analysis and it's uh applications in uh Environmental Health Sciences and other health related areas thank you hello I'm H I'm a faculty member in the department work on uh computational statistics bi infor
matics and cancer genomics uh in particular I'm sort of in charge of the H data science program so if you're interest you know have any question regarding that program feel free to talk to me hi I'm Sebastian zelner I'm a professor in the department my work is primarily in statistical genetics and population genetics I'm also the co- director of Precision Health at the University of Michigan hello everyone um welcome my name is shano I'm a professor in B statistics I mostly work on machine learn
ing for genomics hi I'm Jee Morrison I'm an assistant professor I work on statistical genetics and causal inference um maybe we should get m before we get all the way to the end just I give it all hi everyone hi I'm d uh I'm professor in the department so I work on machine learning pres medicine and the survival Nares as well good morning everyone uh my name is Peter S uh professor of the department and I primarily work in a smart he house uh using where devices to understand digital features to
guide our uh behaviors or nutrition and other aspect of your life style and so I extensively work with people from nutritional Sciences environmental sciences and neology thanks hi I'm Phil bster I'm an associate professor uh I'm interested in data integration problems and applications in cancer and ECMO hi I'm Jeremy Taylor I'm not in your book I'm substituting for Fano so you're looking for fan you have to talk to me sorry um so I'm Jeremy Taylor I work in survival analysis longitudinal data
um missing data a lot of cancer applications and I don't know if anyone did bdsi here in previous years anyway you possibly came to my backyard for a a picnic hi Matt xowski I'm a clinical associate professor in the department I work in statistical genetics population genetics and epigenomics hello everyone I'm MOSI banery I'm research professor in the department my areas of Interest are predictive modeling survival analysis correlated data methods with applications to health policy and outcomes
research and specifically in cancer and pediatric heart disease uh many of the um faces here I mean I'm I'm trying to put uh faces to names I serve as chair of the master admissions committee so if you have questions feel free to come and talk to me hi I'm uh Mike Elliot professor of biostatistics and research professor at the survey Research Center at The Institute for social research uh have a lot of interest in survey statistics uh causal inference longitudinal data um work in a huge variety
of various applications that have some degree of relation to those topics but nice to meet you all oh hi my name is El and I'm working in cancer research nice to meet you everybody here thank you hi am V balad utani I'm a professor of Bio statistics here my research is mostly in basian modeling and machine learning with applications to genomics cancer genomics and imaging I also direct the cancer data science uh unit here on campus with a with a Cancer Center here and Y thanks hi everybody I'm
Mike Banky I'm the longest serving faculty member here in BIO statistics I lead our Center for statistical genetics and genome science training program I am thrilled to see all of you here today welcome and I look forward to talking with some of you during the meetings from 11: to 1 and also at lunch after that I am Nicole fenick I'm not a faculty member I am a staff member I've been here for 17 years and I help with the admissions and recruitment area um I hope that uh you found everything all
right and as we trans transition to the faculty meetings um as we exit we'll be giving you maps to the locations of the different faculty um you also have room numbers on there but I'm fairly certain you don't know what those numbers mean so we have Maps um and we'll be handing them out here so uh and the faculty do you know where you we'll help you get there too okay thank you e e e e e e e e e e e e yeah okay we're good hi everyone um tell you that are online my name is Mike I am a 3year PhD s
tudent in the department um and hi I'm R I'm Hannah I'm also a third-year student in the department for your PhD completed my masters here in April of this year um today we're going to talk to you about student life and our stuff like that and then after that we're going to move to a student panel you guys can ask us questions about our experience here or anything you guys are wondering about in our or Michigan biostats um so to start we are the department of biostatistics um we are within the S
chool of Public Health here at Michigan um and then within that we are in within the college town of an arbor um do you anything to add um yeah Ann Arbor is a really awesome college town to live in it's a really awesome place to be a young person um and the big University of Michigan gets a little bit smaller within the department of biostats and within the school public health as well so it's a great way to make your community a little smaller as you work your way in on this little V diagram ye
ah and as Hannah said uh one of the cool things about an haror is that we are a college town we have a lot of you know young people coming in Young cols from like all the departments across the University of Michigan um so it's a really Lively town for you to live in um next we'll be talking about our department um I'm sure you've seen from the previous presentations that we are a larger Department um currently for this year we have 259 students 124 staff members uh 48 full-time faculty and we h
ave over 2,000 alumni um and then I'll also add that the Michigan Alumni network is really strong so this is a really awesome place to talk to people if you're looking for internships or JS and things like that we just have a lot of really because of this big department we have people going all over the place um so not only while you're in this department but then also people who have gone through it as well are great resources to you uh here on the slide we have a bunch of the student resources
and other Student Activities that are involved with our department um we have a bunch of academic support there's um Library resources Jour resources from Michigan um our department facilitates first year study groups for the student so as you come here um for while you tackle your first year school coursework you'll be doing um study groups uh which are voluntar uh voluntary like you don't necessarily have to do them but I strongly encourage you to do some study groups um it's a great way to m
eet students a great way to understand the course workor better um there are groups that are led by a course facilitator who is a senior PhD student um there's we have a wellness and inclusion Advocate her name is chadima um she enhances and improves the student experience through activities that promote wellbeing and a mul cultural Community um Chima hosts multiculture food and game night which is where students bring food and culture uh and games from their background um and it's just like one
night in the department where everyone gets together and talks about you know what's going on in their life and you can play games eat food and stuff like that uh we have some great Computing support from our department um such as from Mike who's one of our our software developers um he helps students to write our packages um he helps them to you know understand R for the homework and stuff like that um and they he host a workshop and there's just a bunch of software development support and stu
ff like that um Community engagement uh we are engaged in the community I'll let you talk about statcom cool yeah so I am a co-president of statcom so I am biased I think it's really cool but it's a really great program that you can get involved with where you get to um as a grad student you get to work with Community Partners statcom stands for statistics in the community um and it's a great way for you to get your hands on some real world messy data um because the ones you see in your classes
are all pretty curated um and it's a volunteer situation where we create groups of students to work with a community partner um and then you get that real Hands-On data experience with the analysis anything after the data collection phase and we have some really awesome Community Partners we work with people around the ant Arbor Detroit area into Chicago we have some Partners in Northern um Ohio as well um it's a really awesome program a great way to get involved with some real life research typ
e things um and then also with Community engagement we also have lots of other groups you can get involved with one of which is the biostats Student Association or BSA um they're a group of biostat students who put on events throughout the semester um some favorite ones that I've um participated in they have therapy dogs they bring in near finals which is really awesome we've had some um events where we've gone out and gone to the creature Conservancy which is something in an arbor with a bunch
of exotic animals around here which is cool um and lots of ways to just connect with other people in bioa outside of the the work and school environment um and we have other um also ways you can get involved in the department things like peer mentoring committee um curriculum committee and ways you can offer effort to the apartment um in a really cool way to get involved right away oh yeah that's the next thing departmental committees yeah so I'm involved in the peer mentoring committee which is
really great you get to um as a seconde student and Beyond you get to partner with the first year students um as a mentor which is really great and then also that means for you as first years you get a mentor student right away somebody who's gone through exactly what you're going through um and they are a great resource to ask questions about you know which electives might make sense based on your interests how to um you know navigate the transition to grad school even silly things like where
do I go find this kind of food in an arbor so it's a really great resource right away when you start here um in biostat to make that again maybe potentially big seeming Department a little bit smaller right away yeah I myself um I'm on the student Recruitment and the health data science committees uh I'm also a student representative at the faculty meeting so we're a very Democrat uh Democratic dep department where there's like a lot of say from the students and given that we're such a big depar
tment um I think spe specifically our chair AAR has made a considered effort to give the students a voice which I think is really great um lastly on this slide um there's always seminars workshop and faculty and student interactions um again because we're at the big department uh the leadership here has made a considered effort to make sure students and faculty are getting to know each other in smaller settings outside the classroom um so that things like BFF time um there's a weekly seminars wh
ere speakers from all over the world which you can learn more about research and different research areas um the graduate student working group which is where a graduate student will present their research for one night and there's only graduate students allowed in the audience so it's a kind of a cool way to show your research without you know pressure from faculty asking you super hard questions or stuff like that um there's also faculty launches which are once a semester so twice a year um wh
ere you get like an Excel sheet of like all the faculty in the department who signed up to do this um they pick a restaurant and they take a group of like five to six students out to the restaurant and that's a great way to know the faculty on a more personal level and also a great way to have some lunch on the Department's money um given that we are graduate students um we of course are going to advertise the ones that are free food as you might have been wondering why some of them were B old a
nd I toal size and that's why um next we're going to talk about SP um there's we're there's 40 research centers and initiatives within the School of Public Health as well as more than 40 student organizations um there's the career development center and the writing lab which are resources provided by the school public health uh and these are great resources that you can use I know the writing lab um they help with everything from your Capstone course with $66.99 they can help you write your pape
rs for that course they help you with their dissertation um anything like that and then the career center help you with your um cover letters and your resume and stuff like that as well as connecting you to internships and potential jobs um um yeah so this this slide is a little bit more about broadly the University of Michigan and the really awesome things that we have available to us as a bigger institution um so as you can see here on the left um this is a picture of the big house and you kno
w I'm sure you've heard about Michigan football we just won the National Championship it's a really exciting place um and so we have really awesome things to do in the community we have you know lots of sports events we have musical theater we have concerts all sorts of things put on by the institution which is really great um and again um we are a really top rated University we're consistently ranked in the top um just as a university um as a whole but also Ann Arbor is ranked consistently is a
really great place to live so all these these things together really bring a lot of opportunities for you to get involved there's lots of things to do um and beyond that there are also just again with the university we have three different gyms on campus so you have ways to be active um in three different areas there's one that's under renovation really close to SP right now which should be finished I want to say in the next two years or something so lots of really exciting things coming there
um and also we have a lot of really awesome resources things like caps for mental health resources available to students at the University um and our University Health Service is also available to students in really close to it's walking distance yeah as well um again Michigan's a big school um we're a big department and a big school and a big university um so there's a bunch of intermal sports and stuff like that um Michigan has its own quiddit team if you're a fan of Harry Potter um and I do a
lso mention that one of the gyms is completely under renovation as Hannah said they completely knocked it down they're putting up a new one um but instead in the meantime the university put up like this big in this like big insulated tent um on one of the fields as like a temper regy that you can use so we do have three gym facilities um and as long as you have your M card and you're currently enrolled as a student you have access to them yeah so going back to again y Arbor being great to live u
m one thing I'll note you know people ask a lot about the weather um I think the most important thing to have is a warm coat and some waterproof boots because even though you know people talk about the winter I think it's less about um you know the snow it's more about the snow melting so make sure you have some waterproof Foods that's important to have um and also um we are really close to Detroit and also the uh DTW the airport it's only a 30- minute ride which is really easy we have the Michi
gan Flyer bus that is I think $15 to get a ticket to the airport so it's really easy to if you're from farther away to get back home Detroit is a hub it's pretty easy I live um I'm from the West CO G area and it's really easy for me to find a flight home which I really appreciate being further from home um and they also we have an Amtrak station that's really close um to the downtown area as well you can get buses and trains to Chicago pretty easily and at other cities in the surrounding area so
it's a pretty trans um it's a really interconnected city as well yeah I definitely Echo everything Hannah says being you know close to Detroit um airport is really convenient and they have the Michigan Flyer which is a bus you can take for like 10 15 bucks to go there which is amazing um I do want to talk about our public transportation system um so as a University of Michigan student with your M card you have access to not only the Michigan buses but the ant Arbor buses as a whole so you can r
eally get around an arbor and some areas of ipolani um for free with your M card which is really great um so lastly we have these pictures um we really want to emphasize here as students the culture not Department that we have our department really has made a strong effort to build community and our department has so many events that you can attend um here we have the Christmas party we have our fall picnic um typically on Valentine's Day they have um a dumpling making section which you can see
up here um which is really cool we have ice skating in the winter uh when the weather weather gets a little warmer we have our biostat versus Epi cornhole tournament um we are the current Champions um if you want to help us keep that feel free to come here if you go to cornhole uh we have movie night um graduation and of course um we're a strong research institution so before students graduate they put together like a senior research showcase uh where students get a chance to talk about their di
ssertations and you can come learn about them um that's all that we have for our slides now I want to get to the more important part which is answering your questions um which would be happy to do um student about Services I feel that many students are planning to study in the PHD program and I would like to ask ADV about programs care service for those who want to go to work directly because there's no very yeah that's a good question um so the question I'll repeat it is that um what career res
ources are there for people that want to do terminal Masters don't necessarily want to do PhD um so definitely a lot of our students go on to the PHD but we have plenty of students that go on to a job after they do the Master's um I would say for the I would actually maybe say that the Alumni network is maybe even more important than career Resource Services uh we have 2,000 alumni out there that um are familiar with our department they know our department puts students through rigorous training
and they produce good bio statisticians um we have many students that get internships after their first year of their MERS and then they get a job after that typically for the internship some of them just apply themselves and you know we get scooped up really easily because of our machine brand and stuff like that um do you anything want to add to that yeah I'll also say that so Michigan we send students all over the place we have just thinking of our cohort of people who left after the Masters
we have people that went to um Indianapolis to Baltimore to Boston um just the top of my head those are the examples that I have Chicago so we send people all over the place so again that network is really important um and in addition to as well the career services we also get lots of emails again from the alumni who have graduated from this department that say hey we're looking to hire we'd love to have one of your students um so we also just get lots of emails of opportunities as well and it'
s definitely something that I would recommend taking advantage of and all I think of my friends who had internships between their first and second year of the Masters ended up getting jobs with those same companies as well after the Masters was done um so it is a really strong Pipeline and I would recommend taking advantage of it definitely and this question also mentioned um a lack of um official report for that so I mean this is obviously anecdotal evidence but some of the master students that
I know that left they went to things like Veterans Affairs one of them went as a data scientist that read it um Eli Lil is a bay one we have multiple graduates go there each year which is in Indiana um ABV is another one and our TP Master students typically don't have problems getting jobs yeah we also had two people from our cohort go to work at Harvard as a biostatistician as a researcher so again just a wide range of things and Nicole do we have I thought we did have a a report of some kind
maybe I'm well yeah was okay I feel like there might be something online as well so I would I would look I'm pretty sure that Michigan has a pretty good outline of where biostats grads go so I would look into that again because I swear I have seen it before questions first is what are some ways that students have paid or paying for their degree yes that's another really important question um I'll start so obviously if you have a funding offer from our department uh that's great if you don't some
students take out loans kind of like undergrad is the same thing um there's one thing that I do want to mention that maybe you haven't heard of yet in these presentations um it's called rackam cost sharing um so you yourself um can apply for external fellowships um that would be like your own initiative um to apply for them and then if you're awarded one depending on the award that you received um Rockham graduate school can match some of that um funding offer um for example like if you have a
certain offer from the NSF or whatever um and it doesn't completely Co cover your tuition Rockham might cover the other half or if they you have a tuition offer but you don't have health insurance um Ram can provide that for you um so the specific details are online is called The Rock them call sharing but that's another way that students have funded their own degree and I'll also mention that if you don't have a funding offer from biostats as like you know a GSI or gsra you can also look for th
ose in other departments that is something you certainly can do I have friends that found a particular thing I would recommend is teaching statistics classes and non- statistics departments um because oftentimes those students in those departments don't want to teach those classes because those aren't their their areas of Interest so that would be a really good place to start look for the positions or or you can do something in a as a GSI that a graduate student instructor in case that wasn't cl
ear um in an area that you're comfortable with that's not statistic so I had a friend teach a language like a GSI for a language class just because she was a double major coming through and that's just what she thought would work for her so you're not restricted Again by looking just for funding within biostats as well that's the benefit of having you know that big large research instit institution at your disposal another question is I've heard that because the master's program is a little bger
feel competitive for opportunities in just general is that true and could you talk about Department culture environment sure um so the question talks about um competition for resources and St isite given the department size um I would say for the master's program we do have a um competitive nature among the students particularly for spots in the PHD program um I would say if you're a terminal master student um you're obviously there is kind of there's a competitive environment but I would say I
wouldn't necessarily say personally that there's competition for resources outside of that things like the writing L the resources or jobs or internship or stuff like that I wouldn't would you agree with that yeah I would agree with that I think a lot of the competition that um this might be referencing is just people who are interested in getting the PHD after the Masters terminal Master students it doesn't really affect them so much and then people who are direct admit PhD it doesn't really a
ffect them as well um and there's certainly again it's a big research institution so there are lots of resources available to you um so yeah I just taking advantage of those things even just outside this department as well I think a lot of people Overlook those resources as well yeah so the second half of the question about Department culture um I would say interactions with the faculty I've only had positive ones we're on a first name basis which is really nice you know all the faculty are obvi
ously like World round researchers in their field but um I would actually say my favorite aspect of them is like when I just get to meet with them at lunch and stuff and they talk about and they brag about their children stuff like that I think that's one of the most important things that we have in our department um but I do think that the person that has this question talked about the competition I think it's important to emphasize that there are many Master students in our department that are
competing for spots for the PHD program and that's probably where you've heard the competition part come from and it's pretty accurate yeah and and going back to that culture comment too I I think the really cool thing as Mike was alluding to all the professors here are people which I really enjoy and I was really surprised by coming to such a big institution um so the faculty are really awesome to talk to one-on-one and also I would say the students that we attract tend to also be really kind
people focused people I think that's a mixture of the fact that we're a biostats department which is pretty grounded in those applications and people but also I think Michigan just as a place and as an institution um they tend to attract really people focused people as well you know you breed what you um what you put out so I think we do a really good job with that yeah and maybe I should rearticulate there is competition among Master students but there's no like real like none of the master stu
dents here are like nasty about it every like I would say it mostly um encourages like stress among the students it doesn't really like no one here is like mean and stuff like that yeah um why didn't you pick graduate school and is there anything you would change about the department or your experience at Michigan go first yeah sure I can go with this time so um I when I was applying to grad schools I applied quite literally all over the country um so Michigan wasn't particularly on my radar whe
n I first reached out I actually almost didn't apply because I didn't think I would get in um and every interaction that I had with Michigan was incredibly positive so that made a really big difference to me went to a smaller liberal arts school for undergrad and I really wanted to make sure that where I went for grad school I would be supported and Michigan really really came through on that for me um I got a phone call when I was admitted from the person that I mentioned in my essay that I wan
ted to work with saying hi we'd really like to offer you a position in our program like that really meant a lot to me um I also for you know full transparency I was offered a fully funded position which did make a big difference for me but apart from that I compared to the other universities I was really seriously considering Michigan by far in a way made me feel the most valued as a person which was really important to me yeah I Echo a lot of things Hannah just said um personally my research in
terest is statistical genetics um and this is a really strong department for that so that's one of the reasons that I came here um another one of the reasons is again as Hannah said everyone that I interacted with um made me feel supported and made me feel like that like my concerns and questions were important uh and to me that was really important um you know this is a big department so having your small little nicho faculty and students that you know is really important to you know make your
experience to that much better yeah and I'll also say Michigan is a great place to go if you don't know exactly what you want to work on yet because we have such a wide variety of Faculty working on so many different things um I thought I knew what I wanted to do when I got here and I figured out once I got here that wasn't exactly what I was looking for and it was really easy for me to Pivot and start working with somebody else on something that I better enjoyed I think the second half of that
question is what we would change about the department um if I had to pick one thing I would probably I would say our department has kind of a traditional SL old school approach to coursework um I wouldn't say that the coursework itself is like out of date but I would say a lot of the faculty here really emphasize coursework um maybe a little too much I think when you think about your graduate school education you should you know think about like what you want to do when you go there um um I woul
d say you know at some research some places um they train their junior students by getting them involved with research kind of right away and the courses are less important and you're prepared for your future your dissertation you know with that research experience um here I would say the faculty really prepare you for the digitation with their coursework or at least that's their attitude for it um I would say a benefit of that is I would probably presume that the instructional quality here is m
ore important or is better than it is other places just because our facult place a stronger emphasis on coursework in general um but I think our department kind of finding more of a balance between coursework and research especially for junior students would be beneficial uh yeah and one thing I I think I was surprised by coming here I don't know if this is necessarily that I would change or could change but I was a little bit surprised that biostat felt a little bit separated from the School of
Public Health um so there are certainly opportunities for you to get involved in other ways in the department you know like with the um groups that we mentioned earlier that are SP groups rather than biostat groups or Michigan groups um but I was really surprised by how separated we felt as a department um and I do think the department again is a really good support network on its own um but I anticipated there would have been more crossover than there is okay um you have collaboration with the
Department of Statistics how do you see yourself comparing with a St student I can start on this one so um we have a lot of statistics and biostat students taking classes across those two departments so for example I took a beian stats class in stats and I have a friend in stats who is taking causal inference right now with me in biostats so there's a lot of flexibility in you know going back and forth which opens opportunities to take more classes frankly um which I enjoy um in terms of like w
hat you get out of each degree the core classes for both stats and bio stats are cross-listed so the courses teach you effect ly the exact same thing um it's just that you're getting you know different professors maybe teaching these classes and in biostats generally speaking all of our applications are in public health or Healthcare um which really when it comes down to it is really the only big Nuance difference um with the additional piece that maybe um we have a couple classes in biostats th
at stat doesn't have which would be clinical trials and uh survival analysis because those are pretty biostats focused things but otherwise just as a you know looking at those core courses and your outcomes as a grad student afterwards um you can go get exactly the same jobs yeah I would agree with that and I also say our Department's probably more interdisciplinary than theirs in terms of like the collaborations that we have again as Hannah said all of our applications are pretty much in human
health so we have departments so like we have collaborations with like genetics and epidemiology and all the school public health courses in Michigan medicine and the roal cancer center um and I would probably say on average when it comes comes to courses particularly the ones after the core courses the upper electives are probably more applied on average I would assume their courses and for like the doctoral degree and stuff like that is probably more dependent on like measure Theory and theory
and stuff like that we might be a little more applied adding to adding on to the question regarding competition how about competition for certain Labs um I so the questions about competition for certain Labs within the department um I think based off the way that our department funds students um you will be matched to a professor um I don't necessarily know if there's competition among life because you typically don't have like there's not too much fighting for spots because I think the departm
ent kind of decides it for you right yeah and and I'll also say that just because if you want to work with someone and they don't have funding for you you can always work with for credit that is certainly an option um like right now I'm working with my dissertation adviser and she doesn't have funding for me yet so I'm teaching for my funding instead so and I have a friend who in the Masters wasn't funded but still wanted to work with this particular individual and he just took it for credit it'
s like an independent study so there are definitely Avenues where you can work with who you want to you just may not get funding directly from that Source if that's the question about resources yeah and I also mention that if you're interested in the PHD um you can do your dissertation with someone and have them and not be funded by them and you could do your gsra or GSI as Hannah does with someone else um so a faculty have the bandwith to take you on they still can even if they can't necessaril
y fund you so I wouldn't say there's too much competition among Labs right do Master students commonly have opportunities to engage in research projects with faculty members that do not involve a research assistantship yes those have to be sought out they're not going to be just given to you but they are certainly possible um I have friends who sometimes certain courses that you take like I think in particular 6 620 maybe the health Big Data class taught by Peter song I have a couple friends who
did um you do a project in that class all semester and some of them were really excited about what they worked on and so then they just kept working on it and ended up publishing it later so there are certainly opportunities available to you you just have to see them out because they aren't just going to plop in your lap and as you mentioned you can do research for credit right yes and you can do research for credit as an independent study that is certainly an option as well and I'll also go ba
ck to statcom I'll plug that again that's a really great it's not a research opportunity necessarily but it's a great way to again get involved with some messy real life data in a way that research might be doing yeah and as Hannah said these are things you probably have to seek out yourself which is important is all questions um well thank you so much for coming um if you have any questions feel free to reach out to us um we're here to help you have a big decision to make picking your graduate
institution um it's a big financial and time commitment so you deserve all the information you have so my biggest advice here is just don't be afraid to reach out and ask questions even candid questions or contentious questions yeah and you can find both of our email EMS on the admissions ambassadors page for SP as well so that's also feel free to reach out to me afterwards or Mike afterwards um we're happy to talk to you and I I do think that in this decision process in any institution you're a
pplying to I think the most important thing is to talk to students because they will be honest with you so reach out ask those questions um and figure out you know what what's the vibe that fits you really well and I should probably introduce myself I'm Nicole fenick I'm the academic program manager which means I basically lead student services within the department um and I want to thank you for attending and I want to reemphasize please reach out if you have any questions um we will be sending
a survey and in that survey you'll be able to list your name and preferences if you would like to be contacted for further information um and with that um I thank you and I hope we will see you in late August which gets earlier and earlier and go blue thank you
Comments
Go blue and go trump