Andrew Baumgartner, PhD

In this down to earth interview, intern partners, Kalea and Yannell, got to talk with Dr. Baumgartner, a theoretical physicist in the Hadlock Lab at ISB. He discusses various aspects of his job, his hopes for the future, and provides relatable advice for anyone in or aspiring to be in STEM or medicine: 

Can you talk a little bit about what a theoretical physicist does. What was your journey to becoming a theoretical physicist and what inspired you to pursue this as a career? 

It was probably around high school, junior or senior year, when I was really kind of thinking of what to major in in college and what kind of career to go through. I’ve always really loved the stars, looking up at space, and all the constellations. My mom is a science teacher, so growing up she would always kind of hammer that stuff home for me. She really wanted to be an astronomer, so that kind of sparked my interest in science. I started taking more math classes, AP Calculus, AP Physics, and I was really enjoying them. Then I went to undergrad and started studying physics full time. It went well, I luckily had a really supportive and small department at my undergrad, so it really fostered a lot of personal interest in various fields of physics. It all started liking space stuff and then I kept thinking deeper and deeper about astronomy and it all came back to Physics! Then the more I learned about quantum mechanics and string theory I was like “oh this stuff is so cool!”

What a theoretical physicist does in a physics context is use whatever mathematics is at your disposal to understand the mechanics of the universe at a fundamental level. Really, what that means is studying matter and energy in various forms. Those are the two main drivers, everything kind of boils down to that. There’s a number of different types of theoretical physicists too. There’s people who would build mathematical models to describe phenomena that people would observe in a lab (which is more similar to what I’m doing now). In comparison, what I studied in grad school was more abstract physics. It was more using math and manipulation, of not only using calculus and differential equations, but a lot of higher pure mathematics such as differential geometry, topology, and group theory to really gain new insights into the physical world through the mathematics alone rather than using math to explain something that came up in an experiment. There’s kind of a flip side there. 

How do you apply your theoretical physicist background in physical and mathematical principles into your current work with biology and medicine?

Obviously, being good at math is a strong advantage to that. Now I do a lot of statistics and machine learning which is different math than what I was doing before, but it all comes from the same place. Definitely having a strong mathematics education helped me land this job for sure! As far as philosophy for approach to problems, it’s much different than bio. Most biologists and people in medicine study the interacting parts of the body such as how Rna gets translated into protein, what different genes do and how that might affect pathology of certain diseases, but there’s a whole lot going on. Biology and medicine is very very complex and complicated. Physics on the other hand, especially the physics I was doing, while it may sound very hard and complex, the ideas are actually extremely simple. That’s kind of why you can do theoretical physics, because you have a strong set of really simple kinds of baselines to go off of. Then as you go up in length scales, let’s say, going from nuclei and atoms to molecules to biomolecules to DNA, it just gets so so so complicated so fast that those core principles that allow you to do theoretical physics kind of don’t really help you as much anymore. Thus, what I really try to do is bring it back to that. I try to approach problems by looking at the simplest baseline. What are the simplest variables and assumptions that I need to kind of explain what is going on in this problem.  That is really where my physics education kind of seeps into what I’m doing now. I have a number of friends who also study biophysics that I met in grad school, and one of them put it really kind of perfectly in that 

As a physicist with a physics training going into biophysics you kind of learn to filter out all the noise and whatever you don’t necessarily need to solve the problem at hand.

I think that is an invaluable lesson that I learned in my education. 

Can you describe your current research projects in the Hadlock Lab and how they contribute to what is being developed such as new technologies, cutting edge research, etc.?

This is also kind of motivated by my physics background because obviously in physics things move through time. We’re all moving through time. Most of the things you do in physics is time dependent. Through physics, you learn how to study the evolution of systems through time. Really, energy when it boils down to it is fundamental to how we move through time. So where I was coming from, time played an incredibly important role in everything I was doing, but coming to once again biology and specifically the medicine we do here is the time factor. Things are already so complex when you’re dealing with static data and so kind of bringing in time, how systems will evolve over time, how cells will differentiate, how different diseases take hold in your body, the temporal aspect is incredibly difficult to deal with. I’m hoping and kind of what I really want to do, is kind of focus on these temporal aspects and how do you put into context certain longitudinal data and extract time dependence from it. 

Currently, I’m working on a project with Jenn that is doing exactly that. I’m looking at the Electronic Health Records and what I’m trying to do is infer trajectories of vital signs or lab results, so you that you can really see that if you’re hypothetically sick, how your blood pressure changes over time or how your pulse changes over time, and is it different for different diseases, etc. Dealing with the EHR in particular is incredibly difficult because it’s not an easy data set to work with because there’s a lot of missing values due to it’s sporadic nature. In EHR there’s a lot of irregularity in how the temporal aspects of the data is collected which makes it hard to do analysis on. 

One of my projects is attempting to kind of untangle that and give everybody in your data set (all your patients) kind of a similar temporal profile so it’s easier to compare things across the board. I’m also working on another one that deals with Alzehimers disease, which is actually worked with people in the Hood-Price Lab, but I’m also trying to use not so much the same ideas, but the tantamount concept of velocity. There’s a lot of work that’s been done in recent years to kind of bring these ideas of velocity into the biological domain so you can kind of see and visualize how your cells can change and stuff like that. We’re still kind of exploring things and I’m still kind of running some data, but we’re trying to use these ideas of Rna velocity to understand how different cell types are implicated in Alzehimers disease. Those are kind of the two main projects and I’ve also got things in the wings that are kind of half baked ideas that are in various stages of becoming projects for real.

What research do you hope to continue or start in the future of your career?

There’s a ton of different tools, mathematical frameworks, and ways to analyze data that I just want to apply to some sets and see what happens! You know a lot of it boils down to this time dependent concept still. It gets a little more tricky when you’re dealing with populations of people for individual counts of Rna because there is so much fluctuations and variance in your data. It’s so noisy that it’s kind of hard to look at an individual’s profile and really get a sense of a deterministic trajectory. For example, I do something over here and there’s a predefined way of telling what it’s going to do over here in a year or something. That’s really hard because these are such noisy systems. 

Luckily, there are certain frameworks that allow you to instead of studying an individual’s profile, you can study probabilities. You can build models that determine what’s the probability that I’ll be sick, what’s the probability that I’m going to get better, and things like that. From these, you can shift from saying “Okay if I give you this medicine you’re definitely going to get better” to “If I give you medicine, there’s a probability that you’ll get better in two days, the probability increases in four days, and it decreases in eight days” or something like that. It’s actually a pretty big shift from understanding and analyzing systems in the first way versus the second way. So, the projects I want to get going on are kind of more towards that second way and really kind of looking at these probabilities and the ensembles of the patients as a whole to not tell you exactly what’s going to happen, but give you a list of likely outcomes. 

Interns Yannell (upper left hand corner) and Kalea (bottom middle) with Dr. Baumgartner (upper right hand corner).

What is an ugly truth of your research/job and what are some ways that you’ve found help you navigate through that? 

Oh, I can break that down into a couple of different things. There’s ugly truths about data and that is usually, especially EHR, things are not really recorded uniformly, they’re not inputted uniformly, it’s very noisy like I said, there’s a lot of missing data, etc. It is really kind of hard to deal with that stuff. People in the lab and elsewhere have come up with great ways to do that, but once again it kind of makes it difficult for this kind of personalized trajectory approach to deal with. There’s also a lot more software debugging and just general computer nuisances that I didn’t really have to deal with in my former life because I wasn’t doing so much computer work, but that definitely is a learning curve to get used to. For example, how to troubleshoot and debug code and install packages and stuff like that. It can really be a pain in the butt. Part of living in the 21st century that’s nice is that there’s so many options for things, like you have many different kind of platforms that you can run your code on, but what is unfortunate about capitalism is that they’re all competitors, so a lot of there platforms are very different. If you want to use them, you kind of have to relearn all the different features for one platform.

Those are kind of typical nuisances, typical hurdles, and there are of course the environmental hurdles. Everyone at ISB is really great, but part of the ugly truth about being a scientist, especially one of the young scientist in the field, is dealing with imposter syndrome, constantly kind of dealing with the competitiveness of the field, feeling like you’re not good enough, trying to keep up, etc. Those definitely take a toll on your mental health and learning how to compartmentalize those aspects of the job and deal with them in a healthy way is difficult. It’s something that a lot of people have to learn. Luckily, we have a really great group and Jenn is an understanding boss, so there’s not so much external pressure where it’s kind of really crushing down on your mental health, but it’s more like you know “publish or perish” type of deal. That’s kind of always living over your head. If you want a career in this stuff, you’ve got to be grinding, but you know “do you have to be grinding?” I don’t know, maybe you just got to be really efficient and good at your work. It’s definitely hard to strike that balance and it’s something if you’re going into STEM or medicine you’ll have to tackle with down the line. But, I wouldn’t worry about it too much until you have to. Meanwhile just have fun because that’s what life is all about. 

Where do you hope to see your field go in the future/what do you hope to see change over time?

I think the biggest thing I want to see change is definitely more diversity and inclusion in the fields. Historically, access to higher education has been reserved for few privileged individuals and people who generally look like me. That’s not how you foster a good kind of scientific environment. The ideal goal would be to have equal access for everyone and everyone to have a chance to succeed in the fields they want to succeed in while not being hindered by jerks at the top. Aside from that, which is obviously super important, is having a variety of different viewpoints from a variety of different people from different backgrounds. This will help inform important research questions that help solve problems or answer questions that will help different communities that people like me might not necessarily know because we’re not existing within them.

I think it’s definitely for the benefit of humanity as a whole and science as a whole to have as heterogeneous of a scientific population as possible, so that we can just kind of tackle all the important questions from everywhere around the world and provide equal access to everyone.

Andrew Baumgartner, PhD

I think it was a lot worse in physics than it is in bio from my experience in terms of diversity, but I really hope going forward that a lot more energy and resources will be allocated to increase that diversity at the top.That’s kind of the main thing. As far as scientific questions, where the scientific questions are leading, you know, I just hope that important questions are being answered. I don’t really know/think I have enough knowledge at this point to know what those questions are, but yeah. 

How has COVID-19 hampered or enhanced your research/job? 

I actually got this job during COVID-19. I got hired back in November, so you know it was kind of weird being hired to a new job and not having to go onsite somewhere and talk to people in person. It was kind of isolating at first, but at the same time, because of the nature of my work in grad school I was kind of used to it. I came from a small research group and I was my advisor’s only student. It was in a small group and there was some collaboration, but a lot of it involved reading, doing math, writing a little computer code here and there and that was kind of isolating in and of itself. I was kind of used to working alone and working in isolation, so in that regard COVID didn’t really change that much, especially as a theorist or now someone who writes more computer code. I don’t really need to go into the lab to do it, I can just do it on my laptop from anywhere. I guess one thing I really miss is that I used to go to a lot of coffee shops and work. Obviously I haven’t done that in COVID, so that would be a nice thing to do when 4 oclock rolls around in the afternoon and I’m kind of sick sitting at my desk in the office. Having a change of scenery and being surrounded by other people just living their lives is kind of refreshing and it gives me a little more energy and helps me focus. I guess now that places are opening up again I could do that, but you know I’m still kind of iffy. 

What advice would you give to high schoolers or anyone else who may be hoping to pursue a career in healthcare or the STEM field?

I would say, first and foremost, time management is a huge one. This kind of goes back to what I was saying about the fields being very competitive in the first place. As long as you manage your time well during the day and in college, getting all your work done and studying hard and all that stuff, you will be able to have a good balance between your work and life. This is super important because having a life is fun, but it also informs the type of work you do and the questions you answer. If you go out and interact with people, you kind of know what people who aren’t scientists and specialists care about. This is important especially in being able to communicate scientific ideas and stuff to other people!

I would say managing your time well and being able to do both good and hard work and still have a life and have fun is some of the best advice I could give.

Andew Baumgartner, PhD

Sometimes you can’t always do that because you’re under pressure or whatever, but trying as hard as you can is also good. Another one would be don’t compromise who you are for the sake of a career. I think that is important for any career. It’s important to kind of not change who you are and how you operate as a person just so that you can get a job or appease your boss or something and that’s true across the board with any job. People like us, we don’t really consider a research institution to be part of the capitalistic machine but it is in some sense. There is some external spending to benefit corporations, and it’s not worth giving your life away to be a cog in that machine, even if it seems to be more important because it’s medicine or science. I still think having a life is the best thing you can do. Don’t compromise who you are and don’t compromise all your time and within those kinds of restraints just do the best work you can and work as hard as you can.I think you’ll find the more you have a normal life and hangout with friends and family and stuff like that, the harder you will work during the day. Part of it being because you know you compartmentalize that time to be like these 8, 9, ten hours, whatever I’m working today, I’m getting it done. Then I’m going to go home and have fun. I think that’s very important for individual mental health and the work you do as a scientist.

Link to ISB Profile: Andrew Baumgartner, PhD | The Hadlock Lab (