Inferring regulatory signal from genomic data

Abstract

The growth of biological data necessitates development of data mining methods tailored towards understanding complex mechanisms of biological systems. This project focuses on issues related to gene expression regulation, namely: identification key genes from microarray data and analysis of sequencing-based localization of interaction sites of transcription factor (TF) and DNA. In relation to gene expression analysis, we focused on: (i) identifying a minimal gene signature cassette and (ii) identifying primary response genes using time-course expression data. Using high-throughput sequencing-based TF-DNA interaction data, we developed models and formulae to (i) rapidly assess the sequencing adequacy of a given library, (ii) model for ChIP fragment size distribution, (iii) model the signal/noise component and ChIP enrichment quality of a given library, (iv) provide an analytical model of random fragment accumulation, and (v) mitigate the effect of systematic biases arising from aberrant genomic copy number of the underlying biological model system.