Search
School of Informatics and Computing Menu

Ramana Dauuluri, Ph.D.: Mammalian genomes are too complex – It is time to move away from simple gene-centric approaches

Watch

Abstract

Mammalian development is an intricate process regulated by multiple gene isoforms and their epigenetic states. Recently, we and others have discovered wide-spread use of alternative promoters and alternative splicing in mammalian genes in various tissues, developmental stages and cell-lines. We built genome-wide inventory of transcript variants, their promoters and histone modification states during normal development, using integrative Next Generation Sequencing (NGS) and bioinformatics approaches in mouse cerebellum as model system (Pal et al, 2011, Genome Research). While the recent advances in NGS are facilitating the goal of studying gene regulation at isoform-level, there are a number of informatics challenges and difficulties that need to be addressed to improve the current state and fulfill the promise of studying gene regulation at gene isoform-level. We are developing novel computational methods that can accurately integrate genetic and epigenetic signatures, derived using multiple experimental platforms. For example, Mammalian Promoter Database (MPromDb; http://mpromdb.wistar.upenn.edu) provides an integrated resource for mammalian transcriptional regulation and epigenetics (Gupta et al. 2011, Nucleic Acids Res). I will present some of the recent approaches developed by our group, with an emphasis on the how those methods have led to the development of a diagnostic assay for molecular sub-typing of cancer patients (Pal et al. 2014, Nucleic Acids Res; Zhang et al. 2013, Genome Medicine). In particular, I will challenge the use of basic gene-centric approaches in biomedical research and argue that one should go beyond simple gene-based analyses but also consider isoform-level information that include gene expression of splice-variants. Looking forward, I will discuss the integrative application of different statistical and data-mining approaches to derive platform-independent classification models for identification of isoform-level genetic and epigenetic signatures.

About Ramana V. Davuluri

Dr. Davuluri has worked in the areas of Computational Biology and Bioinformatics since 1998. His work has focused almost exclusively on promoter annotation and transcriptional regulation in mammalian organisms. His research findings have been published in notable peer-reviewed journals, while the bioinformatics and translational research communities utilize the computational tools (e.g. FirstEF, MPromDb) that he has developed.