RESEARCH SUPPORTED BY THE BIOINFORMATICS IGERT
Examples from recent student research
Research in molecular biology is shifting fundamentally towards the study of complex, multi-component systems, or networks, that underlie the living cell. These networks are described and modeled in terms of their components, component interactions, regulatory properties, sub-networks or pathways, and system dynamics in steady-state and in response to perturbations. Areas where networks research is becoming increasingly important are 1) biochemical pathways of metabolism, 2) interactions of proteins with DNA to regulate gene transcription, and 3) signaling pathways for cellular response to hormones and other molecules which modify activity and control early development.
The networks approach has developed in conjunction with the shift to high-throughput data generating techniques (microarrays, chromatin immunoprecipitation, two-hybrid studies, mass spectrometry, etc.). As a result, it is necessarily dependent on the development of advanced computational and mathematical tools to pool, organize, and analyze vast quantities of data. Networks in the abstract have mathematical properties that allow inferences about their connections, missing components, and dynamics. High-throughput data combined with comparative genomics supply the foundation for these inferences as they apply to biological networks.
The BU Bioinformatics IGERT supports training and research in biological networks. This can be divided into three interrelated sub-areas: regulatory networks, metabolic networks, and protein-protein interaction networks. Across the sub-areas we have leading experts in computational and mathematical methods and outstanding experimentalists studying important underlying biological functions. Strong collaborations, which combine essential biological motivation and validation with theoretical work, exist within and across the sub-areas.
The following is a brief outline of the major research areas sponsored by the IGERT.
- Regulatory Networks. We are studying the interaction
of intracellular signaling pathways with the association and binding of
genes by transcription factors (TFs). Understanding and identifying these
associations are key to elucidating the action of fundamental regulatory
networks. Goals include developing and improving methods for identifying
gene-TF interactions, predicting TF binding sites, and constructing regulatory
network models from data about gene-TF interactions. Beyond the high-interest
specifics of TF binding sites, other motifs and repeats in nucleotide sequences,
collectively, provide “points of action” for modifying gene
expression at the levels of transcription, translation, and post-translation
and therefore have a broad and fundamental impact on regulatory networks.
Examples in DNA include nucleosome positioning signals and inverted, tandem
and other repeats that influence deoxycytidine methylation and the formation
of heterochromatin. Examples in RNA include cis-elements for mRNA localization
and signals for microRNA targeting and suppression of mRNAs. Computational
techniques for motif/repeat identification and study are being developed.
- Metabolic Networks. We are developing and studying dynamical
and steady state (flux balance) mathematical models of metabolic networks
to determine how pathways respond to genetic (gene deletion) and environmental
(nutrient shift) perturbations in order to understand both physiological
regulation and evolutionary adaptation of metabolism. Mathematical methods
are being developed and coordinated with experimental studies.
- Protein-Protein Interaction Networks. We are inferring and studying protein interaction networks using comparative genomics, co-expression clustering based on high-throughput microarray data, and protein-protein docking among other techniques. Goals include annotations of protein function and pathway assignments.