From sequence graphs to processing bias: towards more informative microbiome representations

February 24, 2025

Location:

In person: MIT 26 Room 168

Zoom Alternative: https://mit.zoom.us/j/91620777596

MIT Microbiome Club Academic Seminar Series

Professor Tal Korem
Department of Systems Biology
Columbia University

Monday, February 24th, 4 – 5 pm

Summary: In recent years microbiome data has grown substantially in scale, with applications ever closer to clinical translation. This complexity necessitates new methods that facilitate improved representation and robust data analysis. I will present two of our recent efforts in this area. In the first, we propose a new infrastructure for comparative metagenomic analysis, using a new hybrid co-assembly approach to generate sequence graphs that represent multiple microbes across many samples. In the second, we developed a bespoke method for correction of processing bias that enables improved microbiome-based prediction of diverse phenotypes.

Bio: Tal Korem’s research program focuses on the development of computational methods that identify and interpret host-microbiome interactions in various clinical setting. The ultimate goal of his research is to translate microbiome findings to clinical care, with microbiome-based therapeutics and microbiome-informed clinical practices. He has developed several approaches for microbiome data analysis, inferring microbial growth rates, structural variants, contamination and experimental bias; and has applied these methods in diverse clinical and biological investigations, most notably for personalization of dietary treatment and prediction of adverse pregnancy outcomes. He is a member of Columbia’s Program for Mathematical Genomics (PMG), an Assistant Professor in the Departments of Systems Biology and Obstetrics & Gynecology, and was previously a CIFAR-Azrieli global scholar by the Canadian Institute for Advanced Research.