Q&A with Lydia Owens

Lydia Owens is a residential NCSSM student from Raleigh, North Carolina. She took part in the 2022 Research in Computational Science program during the academic year and Summer Research Experience in Computational Science program, a three-week intensive research experience during NCSSM’s Summer Research & Innovation Programs. 

What is computational research? 

Computational research is basically taking computation skills, like a bit of computer science and a bit of data science, and applying them to other problems in other fields. 

It’s really interdisciplinary by nature, and so not just with my project but with a lot of people’s projects in Research in Computational Science, we’re looking at things like solving Alzheimer’s disease or sports plays and how people make decisions, so there are a lot of different applications in biology, psychology and other fields. 

I chose epidemiology because that was something I was interested in from personal experience but I think the great thing about participating in Computational Science here is that once you learn the basics of how to code, how to build a model, how to analyze data, then you can apply it to any problem you want, and so the possibilities are really endless in that regard. So we’ve had a lot of interesting projects come out as a result of that. 

What has your experience been with SRIP Computational Science?

My experience with the summer computational science program has been a little bit different than what I was doing during the school year; I was doing mostly the theoretical framework of my proposal and understanding the field that I am researching, which is computational epidemiology. 

But in the summer, I actually got to start building some code, trying to get my model to work, and working with my mentor to better understand how I could implement some of the ideas that I had in my research proposal. 

Tell me more about your project: 

My project is basically applying agent-based modeling, which is a type of probabilistic modeling, to COVID-19 outbreaks within prisons. 

I am currently focusing on North Carolinian prisons, so I am hoping to use a data set from the NC Dept of Health and Human Services and understand how policies such as masking, testing, and other events that happened during the first two years of the COVID-19 pandemic impacted the transmission of COVID-19 within NC prisons. 

What made you want to do this project? 

When I was growing up I was a biology kid. I thought I was going to do Research in Biology, but computational science was interesting to me because I could look at more population-based, or larger-scale issues.

I definitely got more involved in those and social justice with the COVID-19 epidemic in terms of learning more about activism and how I can help in my community and with issues in my communities. 

One of the things I had the opportunity to do was intern at a law firm in Washington, D.C., where I was working on a brief to get early release for prisoners due to COVID. 

That’s when I first got introduced to the topic, and so from learning their stories and thinking about it on a personal and individual level, I was able to create this computational project that looks at population groups of other people like them that have similar stories and still need help dealing with the pandemic in jail. 

What do your data sets look like? 

So the data set that I currently have now, I compiled one from the report from the state. I have cases and deaths for both incarcerated people and for staff, but what I am hoping to get from the Department of Health and Human Services is a de-identified data set where I can look at individual prisoners, their moving patterns, and then, of course, their case information and case history, so that will help me create a more detailed model. 

What do you see as the goal of your research? 

It could definitely be a predictor in the future because as COVID-19 goes on in different variants, all you have to do is change the parameters of the model, so it would still be applicable.

A lot of jails don’t have good data collection or good analysis on what exactly is going on with the incarcerated people, and so hopefully this project would lead to better policy, better health care, and other long-term effects outside of COVID. 

Can this be applied to other diseases or viruses? 

I definitely think it can be applied to other diseases. In the reading that I did for my literature review, the main two diseases that were being modeled were COVID-19 and HIV, and HIV is something that is ongoing and will probably have a longer effect than COVID in jails overall. 

So using the same idea as the contact networks and the infection rates, as long as you have a jail model, that can be done. 

Generally just being able to replicate people in their moving patterns has been the main issue with modeling jails. But once that’s completed then you can pretty much overlay any transmission model or any biological model for most diseases that are communicable. 

What do you like about having the summer program to be able to focus on this? 

It’s amazing, because during the school year with Research in Computational Science, we also had other classes and there’s just so much going on. 

During the summer, we are really able to home in and focus on our stuff, and I think I learned more just from having the time to sit around and read and think for a moment and take a step back from my project than if I was constantly trying to meet deadlines during the school year. 

It’s very helpful and also having flexibility in terms of meeting with my mentor and meeting with Mr. Gotwals has also been helpful. 

My mentor is at Harvard University. He’s at the Center for Decision Sciences in the School of Public Health, and so I just emailed him and met him that way.

What have you learned so far? 

I’ve definitely learned a lot more about how complex computational epidemiology is. 

I was able to be introduced to the field by Mr. Gotwals and from my mentor, but specifically looking at probabilistic modeling and agent-based modeling has been really cool because it’s an emerging field in the sense that it deviates from the traditional compartmental model with susceptible-exposed, infected-recovered, and it thinks about graph theory and other mathematical concepts. 

I didn’t originally think I would be interested in that; I was more interested in statistics and data science, but learning this other side and applying what I do know about statistics and other things to this new form of modeling has been really cool. 

What do you think students who are thinking of applying should know? 

They should absolutely know that even if you don’t have coding experience or you don’t think you are a huge computer science person, computational science is still really worth it.