Computational Biology Colloquium | Harmen Bussemaker, Ph.D.

Harmen Bussemaker, Ph.D.
Professor, Columbia University, Department of Biological Sciences and Department of Systems Biology
Lab Website

From millions of DNA reads to mechanistic insight into transcription factor function

Tuesday, December 11, 2018
2 PM
RRI 101

Abstract: In this talk, we discuss how principled biophysical and statistical modeling of deep-sequencing-based functional genomics data can yield unprecedented mechanistic insight into transcription factor function. No Read Left Behind (NRLB), our new feature-based maximum likelihood algorithm for analyzing SELEX data, allows us to quantify the binding specificity of transcription factor complexes almost perfectly over a >100-fold affinity range and an unlimited binding site footprint; it accurately predict changes in gene expression levels in fly embryos when ultra-low-affinity Hox binding sites in enhancers are mutated. NRLB binding models are mechanism-agnostic, but can be examined for signatures of DNA shape readout using a new statistical methodology that we developed. An extension of SELEX that uses barcoded mixtures of methylated and unmethylated DNA ligands reveals that CpG methylation can affect binding by human Hox complexes either positively or negatively, depending on exactly where the CpG is located relative to the binding interface; binding by the p53 tetramer can be stabilized by cytosine methylation both in vitro and in vivo. Finally, we demonstrate how comprehensive integrative analysis of gene regulatory networks driving aging and longevity implicates an unknown zinc finger protein as a key antagonist of FoxO3a, and show that siRNA knockdown of this transcription factor in human cells leads to significantly increased nuclear localization of FoxO3a.

Host:  Remo Rohs