Description of Research Expertise

Dr. Stoeckert's laboratory, the Computational Biology and Informatics Laboratory (CBIL), uses algorithmic and database approaches to annotate genomes, integrate functional genomics data, and analyze expression data in order to generate models of regulatory and functional interaction networks.

Genome annotation and resources:

We have developed a Genomics Unified Schema (GUS) that is used to integrate sequence, sequence annotation, and the results of functional genomics experiments. An important component of GUS, is the RNA Abundance Database (RAD) used for microarray and SAGE-based experiments. We assemble ESTs and mRNAs into DoTS transcripts and DoTS genes to organize transcripts associated with known genes and discover novel genes. Mouse and human DoTS are available at Allgenes (http://www.allgenes.org) and used to annotate mouse chromosome 5 in collaboration with the Bucan lab. DoTS have been used to design PanChips, microarrays for pancreatic genes, in collaboration with the Kaestner lab. The EPConDB site (http://www.cbil.upenn.edu/EPConDB) uses DoTS, the PancChips, and other microarray platforms to provide expression profiles and information on pancreatic genes in support of the NIDDK Beta Cell Biology Consortium.

A major application and driver of GUS development is PlasmoDB (http://plasmodb.org), the Plasmodium genome resource for the study of malarial parasites. PlasmoDB is part of an NIH/NIAID funded Bioinformatics Resource Center to provide Apicomplexan Database Resources (http://apidb.org) in collaboration with the Roos lab (Biology, U Penn) and the Kissinger lab (U Georgia).

Regulatory and functional interaction networks:

Research on how gene expression is defined by promoter and other regulatory sequences is being conducted with a variety of approaches and gene expression datasets. PaGE is a tool utilizing permutations and false discovery rates to attach descriptive, dependable, and easily interpretable expression patterns to genes across multiple conditions, each represented by a set of replicated microarray experiments. TESS (Transcription Element Search System) identifies transcription factor binding sites (TFBS) and recent efforts are directed toward learning collections of TFBSs that discriminate between genes expressed in a specific tissue and all other tissues. COGRIM (Clustering of Genes into Regulons using Integrated Modeling) integrates heterogeneous biological data (e.g. expression data, ChIP binding data) in a principled and robust fashion to identify functional transcription targets and better understand the combinatorial nature of regulatory control.

Tools have been developed to utilize evolutionary relationships toward understanding uncharacterized proteins. OrthoMCL uses a graph approach to group proteins based on species-normalized sequence similarity measured in an all-against-all BLAST search of complete proteome sequences. In another project, phylogenetic profiles are being combined with expression data to generate functional interaction networks. These approaches are being used to annotate functions for hypothetical proteins in malarial genomes especially putative transcriptional regulators.

Knowledge representation and functional genomics standards:

The Microarray Gene Expression Data (MGED) society (http://www.mged.org) aims to facilitate the sharing of microarray data generated by functional genomics and proteomics experiments. Through our involvement in MGED, we have participated in creating the MIAME, MAGE, and MGED Ontology standards and are working with others to extend standards more broadly for a functional genomics object model (FuGE-OM) and ontology (FuGO). We are also involved in a related international effort to coordinate development and use of standards and ontologies for functional genomics (SOFG, http://www.sofg.org).