Model-Based Inference of Transcriptional Regulatory Mechanisms from DNA Microarray Data.

The development of DNA microarray technology has made it possible to monitor the mRNA abundance of all genes simultaneously (the transcriptome) for a variety of cellular conditions. In addition, microarray-based genomewide measurements of promoter occupancy (the occupome) are now available for an increasing number of transcription factors. With this data and the complete genome sequence of many important organisms, it is becoming possible to quantitatively model the molecular computation performed at each promoter, which has as input the nuclear concentration of the active form of various regulatory proteins (the regulome) and as output a transcription rate, which in turn determines mRNA abundance. In this chapter, we describe how our group has used multivariate linear regression methods to: (i) discover cis-regulatory elements in upstream regulatory regions in an unbiased manner; (ii) infer a regulatory activity profile across conditions for each transcription factor; and (iii) determine whether the mRNA expression level of a gene whose promoter is occupied by a particular transcription factor is truly regulated by that factor, through integrated modeling of expression and promoter occupancy data. Together, these results show model-based analysis of functional genomics data to be a versatile conceptual and practical framework for the elucidation of regulatory circuitry, and a powerful alternative to the currently prevalent clustering-based methods. [ABSTRACT FROM AUTHOR]