Assistant Professor
Department of Physics & Astronomy, Rutgers University
Faculty Member, BioMaPS Institute for Quantitative Biology
136 Frelinghuysen Rd.
Piscataway, NJ 08854-8019 U.S.A.
Office: Hill Center, 279
Phone: 732-445-1387
Fax: 732-445-5958
Email: morozov at physics dot rutgers dot edu
Research Description:
My previous (prior to 2004) research efforts focused on developing methods and algorithms for predicting protein structures from amino acid sequences, predicting kinetics of protein folding, analyzing mechanisms of molecular recognition, and predicting binding affinities and specificities of protein-protein interactions. The goal of my current (2004-) research is to predict gene regulation on a whole-genome scale, including the variation that is influenced by cell type, environmental signals, developmental stage, and disease state. Broadly speaking, I seek to improve our current understanding of the "transcriptional and post-transcriptional regulatory code" that links the DNA sequence with gene expression levels. To this end, I and the members of my lab are currently pursuing several projects:- Studies of chromatin structure and its effect on gene regulation (including both individual nucleosome positions and higher-order chromatin structures). In eukaryotic genomes, nucleosomes (histone-DNA complexes) function to compact DNA and to regulate access to it both by simple physical occlusion and by providing the substrate for numerous covalent epigenetic tags. We use sequence-dependent DNA mechanics models as well as techniques borrowed from statistical mechanics and polymer physics to predict nucleosome positions both in vitro (where they are determined by the DNA sequence alone) and in vivo, where chromatin structure is dynamically modified via competition with other DNA-binding factors and the action of chromatin remodeling enzymes. Our computational models are guided by high-throughput measurements of nucleosome positions.
- Biophysical models of protein evolution. Amino acid substitutions that have accumulated in homologous proteins in the course of evolution provide important clues to their stability and function. We are developing descriptions of protein energetics in which observed correlations between amino acid mutations at different positions are ascribed to both protein energetics (e.g. physical coupling between two amino acids in the protein molecule) and compensatory evolutionary effects.
- Prediction of protein-DNA energetics (amino acid - DNA basepair recognition code) using a combination of protein-DNA structural data and high-throughput assays designed to probe protein-DNA interactions on the genomic scale.