Presenter: Eugene Demidenko, PhD, Professor of Biomedical Data Science, Geisel School of Medicine

Host: Michael Whitfield, PhD

The presentation will take place both in person at DHMC, Auditorium H (and via Zoom) and we encourage as many of you to attend in person as possible.

A light lunch will be provided on a first-come, first-served basis.

Please write to biomedical.data.science@dartmouth.edu to request the Zoom login details.

Please see description below for more details.

Please invite your fellow faculty colleagues, research associates, graduate trainees, and post-docs!

Presentation Summary
Visualization of large genomic datasets is of paramount importance for formulating scientific hypotheses before deploying complex data science algorithms to uncover hidden structures, such as clusters or nonlinearities. Novel visualization techniques for big data are demonstrated by the distribution of single-nucleotide polymorphisms (SNPs) in genome-wide association studies (GWAS). To ‘see’ the data in four- or five-dimensional space, we apply slicing to the 3D contour manifold of a kernel density in an interactive setting or automatic animation. Novel methods for correlation-preserving projection of multidimensional genomic data and the gene connection network onto a 2D or 3D sphere are presented. An interactive gene hub visualization of the Gene Planet will be developed, allowing users to zoom in or out to view gene connections at varying levels of detail, similar to popular online software such as Google Maps or Google Earth. Finally, we discuss how sound can be integrated with our visualization algorithms to sonify the unseen data features. The ultimate goal is to equip genetic researchers with next-generation exploratory data research tools that incorporate novel interactive graphics and data sonification to address the complexity of genetic data through controlled data-science video gaming.

Biography
Dr. Eugene Demidenko is a Professor of Biomedical Data Science at Geisel School of Medicine. He has worked at Dartmouth for more than 30 years and holds joint appointments at the Department of Mathematics and the Thayer School of Engineering. He received rigorous mathematical training in probability theory and statistics. Dr. Demidenko’s research covers a wide range of statistical and nonconvex optimization problems and their applications to imaging, shape analysis, tumor regrowth, synergy, ill-posed inverse problems, and other areas. He is the author of three books on statistics. Dr. Demidenko’s book, Advanced Statistics with Applications in R, received the “Outstanding Book Award” from the American Statistical Association in 2022. According to a database compiled by Stanford University, Dr. Demidenko ranks among the top 2% of scientists worldwide based on his publication citation record.