Evolution, prediction and structural organization of regulatory networks in prokaryotes and eukaryotes. This has been a long standing interest (please see publications page for some contributions in this direction) and we would like to extend our current understanding by integrating new kinds of regulatory molecules like non-coding RNAs, RNA-binding proteins etc. into this framework though principally focusing on systems-wide understanding of the design principles and constraints. Early years in genomic revolution focused on understanding individual levels of gene regulation (either transcriptional, post-transcriptional, post-translational etc) and with the availability of high-throughput technologies it is now possible to understand biological systems at a scale greater than individual processes and their complex interplay with other processes (and sub-processes). We wish to exploit massive explosion in post-genomic technologies (such as CLIP-seq, CHIP-seq, 3D confirmation capture etc) and integrate the resulting datasets to improve our understanding of the design principles and constraints on various cellular processes. To address these long term objectives, we will developing novel approaches to mine the data and to build new conceptual frameworks using computational procedures.
Applying network-based approaches in understanding disease biology and in drug discovery settings. This research direction aims to integrate currently available drug-target relationships with in the context of post-genomic data to address a range of questions to understand how drug specificity is achieved in specific populations and their implications for designing personalized medicines. For instance, current work involves using small molecule interactions in humans to 1) uncover the principles for efficient development and delivery of small molecules for different classes of drug and targets with in the context of polyporphisms and copy numbers and 2) understand the design principles of promiscous drugs which are known to have promising therapeutic outcomes and to gain deeper insights into what makes natural products so powerful. A related set of problems being addressed using graph theoretical approaches is to understand the interplay between species in the human microbiome and how their composition varies in different parts of the human body.
Exploiting genomic data for gene function prediction is an area where we want to push the limits by extending the current computational techniques and developing novel ones for mining metagenomes to faciliate the progress in elucidating functional roles for new protein families. Avalanche in genomic data has resulted in the discovery of several new protein families however our understanding of their functional roles any level is very limited. It is also evident from years of research that traditional (single gene/protein) methods for function elucidation although powerful and provide detailed understanding of cellular roles, provide limited coverage of the protein space. So we wish to employ the power of sequences to assign broad functional categories (middle level but accurate) functions to several novel families identified in metagenomic surveys. Such an approach would increase the pace at which protein function prediction can be performed (extending beyond core well-studied protein families) . This area would involve the development of several novel tools and methods for mining sequence data which are complementary to existing methods.
Mining genomes for novel antibiotics by harnessing the power of metagenomics. Here we plan to use both computational and experimental methods to not only identify novel gene clusters from genome sequences which have the potential to encode for small molecules capable of inhibiting bacterial growth but to also employ an integrated approach to understand principles which decrease the rate of antibiotic resistance. For instance, we are currently focusing on PKS, NRPKS and other some well studied groups of secondary metabolic pathways to 1) understand of the evolutionary dynamics of secondary metabolic pathways in genomic/metagenomic sequences, 2) to establish the biosynthetic potential for antibiotic production in different environments and 3) to discover novel members of these well studied groups. This research area is very closely tied with developments in the previous theme on function prediction.