News & Events


Featured Researcher: Jordan Rozum

Posted on September 15, 2021

Cracking the code of whole genome-scale models could lead to cutting-edge medical treatments for diseases like cancer and gain new insights into biology and health. However, researchers face computational problems probing the whole genome models because of their incredible complexity. Jordan is hoping to find ways to reduce this vast sea of possibilities to help scientists better understand the genome.  All of this requires a lot of computing power, and that is where ICDS comes in.

How did you get into this research field?  

I used to study differential geometry and general relativity, and I got my master’s degree in mathematics in that field. I made the switch to biological physics while I was a student at the University of Illinois at Urbana Champaign. I was exploring ways to analyze biological data that involve thinking of datapoints embedded in a curved space. While investigating these possibilities, I came across the field of network biophysics, and was hooked. I eventually transferred here, to Penn State, where I have been studying networks ever since. 

What do you hope to accomplish with your research?  

The human genome has around 20,000 protein-coding genes. Their patterns of activity and inactivity govern the form and function of the cells in your body, and our goal is to better understand these patterns. How do they arise? How can we manipulate them to help cure or treat complex diseases like cancer? These are big questions, and they can only be tackled when many researchers approach them from many different angles. Our approach is to study gene regulation as a dynamics problem constrained by a network of known interactions between genes. Specifically, we are interested in identifying the long-term behaviors (i.e., the stable phenotypes) of these systems and in figuring out how to select one phenotype over another using, for example, pharmaceutical drugs. 

How does supercomputing enable your research? 

We often study small pieces of the gene regulatory network that governs all cells, focusing our attention on regulatory circuits that play key roles in specific biological processes, like cell division, programmed cell death, or cancer metastasis. Despite this narrow focus, we often need to consider several dozen genes at once in a given model. Larger models can have hundreds or even thousands of variables. Even if we assume that a gene can only be in one of two states – active or inactive – the number of possible configurations is literally astronomical. With only a few hundred variables, there are already more configurations than there are atoms in the observable universe. The number of configurations grows exponentially with the number of variables. Recently, we have been interested in whole-genome-scale models, which contain tens of thousands of variables. Identifying the stable long-term behaviors within this vast sea of possibilities requires careful reasoning to reduce the search space and a lot of computing power to sift through what remains. 

What is your academic background? 

I got my bachelor’s degree in math and physics and my master’s degree in math from Utah State University. I studied astrophysics and the mathematics of general relativity. A brief detour into applications of the math of curved space-time to bioinformatics eventually became a complete shift in my focus. I am now nearing completion of my PhD in physics here at Penn State where I study the mathematics of network models of biological processes. 

What important advances do you think we’ll see in your field in five years? Ten? Twenty? 

The process of developing models of cancer and other complex diseases on a cellular level is becoming increasingly streamlined. At the same time, people across the world are already working on ways to quickly build personalized, multi-scale models of complex diseases. Together with the increasing availability of genetic testing, this has the potential to radically change how potential treatments for various cancers are explored, not just at the R&D level at big pharmaceutical companies, but in the doctor’s office. I expect that over the next several years and decades modeling and genetic testing will play an increasingly important role in developing personalized treatment plans for individual patients with cancer or other diseases. 


Related Posts