By Leo Barolo
Kwangmoon Park was always interested in applying research to real-world problems. As an undergraduate student in South Korea, Park pursued a double major in economics and applied statistics to improve his proficiency in econometric tools. These tools seemed more closely related to real-world applications than other topics in the field of economics. But as he delved deeper into his studies, he was increasingly motivated by the application power of statistics.
“I understood that statistics could play a crucial role in any field I could imagine, and I didn’t want to limit myself to its applications in economics alone,” he remembers.
This interest led him to enter the Statistics PhD program, where he received an email from biostatistics professor Dr. Sündüz Keleş. As he read it, Park was introduced to the world of biostatistics and genomic applications. Park is currently a 5th year PhD student at the Department of Statistics and a member of the Keleş group. His research includes studying how genes are regulated by distal regions in the genome, particularly by functional non-coding regions. He is developing statistical tools for analyzing high-dimensional genomic data, including Hi-C and HiChIP, and for linking diverse types of genomic or epigenomic data with better statistical interpretation. Recently, he has focused on 3D genomics and epigenomics data analysis and just published an innovative paper for the analysis of 3D chromatin data.
During his time at UW-Madison, Park has received multiple student paper awards, such as an ENAR 2023 Spring Distinguished Student Paper Award and a 2023 ASA Student Paper Award in the Section on Statistics in Genomics and Genetics (SGG). We asked him several questions about his research and trajectory.
You recently published a series of award-winning papers on the analysis of 3D chromatin data. What is innovative about these studies?
My three recent papers on analyzing single-cell (sc) 3D chromatin conformation data include: 1) A study on simultaneously summarizing scHi-C and DNA methylation measured in the same cell using a novel joint tensor factorization method. 2) Research discovering associations between DNA-methylation level changes and physical interactions between genomic loci through a new tensor regression method with dimension reduction for high-dimensional datasets. The most recent paper is 3) Inferring significant multi-way interactions among (more than two) genomic loci based on scHi-C data. This involves empirical Bayes framework with a clique level summarization of pairwise interaction p-values of scHi-C, which can significantly reduce the multiple-testing burden for researchers investigating SNP-SNP interaction effects in molecular QTL studies.
What is the single person, event, or experience that most influenced your trajectory to where you are today?
My advisor Sündüz Keleş. I majored in economics for my undergraduate studies and initially did not envision myself working in the genomics field before meeting Sündüz at UW-Madison. In my early PhD years, I had a lot of freedom and time to explore the statistical genomics field, and she always provided enormous support for me to enjoy specific data or methodologies that I showed interest in. After several years of such experiences, I became more self-motivated and ready to be an independent researcher.
What advice would you give a young person interested in graduate school or research?
If I were to offer advice to students interested in graduate school or research, I would encourage them not to get too frustrated when things don’t go as they planned or imagined. Instead, I would emphasize the importance of identifying the specific bottlenecks they encounter. In my experience, those frustrating moments are an excellent opportunity to reorganize and rethink everything, often leading to a breakthrough.
Park plans to stay in academia. Despite his recent focus on 3D data analysis, Park is open to any biological questions involving interesting statistical challenges or insights, particularly those related to high dimensionality.