Lab members listed in alphabethical order, with links to our people page.

Sourav Bandyopadhyay
The majority of my research has been focused on the integration of physical protein interactions across different species and has now moved toward the integration of genetic and physical protein networks within the same species. I have focused on methods for cross-species network comparison including the analysis of gene expression changes during HIV-1 latency and reactivation through the use of an integrated HIV-human protein-protein interaction map. Additionally, we have used this type of network integration to implicate several protein complexes and pathways in the HIV infection pathway based on siRNA screening for HIV-host factors.

During the last year, I have worked on the analysis of a large-scale yeast-two hybrid protein interaction network centered on human MAPK proteins. Using a systematic interrogation of physical interactions we have been able to organize kinase interactions into signaling scaffolds as well as into a number of signaling pathways conserved between yeast and man. We are currently in the process of validation of functional roles predicted from this analysis using a combination of co-immunoprecipitation and siRNA knockdowns.

Beyond physical interactions, genetic interactions can also offer significant information about pathway organization. Genetic interactions represent functional relationships between genes, in which the phenotypic effect of one gene is modified by another. We are currently profiling thousands of gene double deletions in budding yeast to uncover genetic interactions through the use of SGA and E-MAP technology. We have also developed methods for the integration of protein complexes and this genetic interaction data to create large maps of the functional associations between macromolecular complexes in the cell. Lastly, we have developed methods for the systematic comparison to genetic interactions between budding and fission yeasts and have uncovered both substantial conservation as well as significant rewiring of genetic interactions between protein complexes.

Jason Chan
Retrotransposons are highly homologous repetitive elements spread throughout the genome that may play a role in the development of gross chromosomal rearrangements. Gross chromosomal rearrangements are one of the hallmarks of the chromosomal instability phenotype of cancer. My research aims to identify what role retrotransposons play in the development of gross chromosomal rearrangements in budding yeast, and in doing so, what role they might play in the development of chromosomal instability cancers.

Han-Yu Chuang
Network-based modeling and diagnosis of cancer development and progression
I am working on developing large-scale, computer-aided platforms of identifying pathway models that lend insight into the mechanisms of disease and may serve as powerful diagnostic markers. My thesis project is focused on the integration of various high-throughput expression profiles (genomic microarrays and proteomic mass spectrometry) with molecular interaction networks (protein-protein and protein-DNA) in modeling cancer development and progression. Our lab has demonstrated successful initiatives in incorporating protein pathways into gene-expression based classification for various types of cancers. My current focus is on the aggressiveness of cancer cells in disease progression of chronic lymphocytic leukemia (CLL). Besides sorting out the underlying mechanism of CLL progression, I will also develop better algorithms in predicting aggressive potential of newly diagnosed patients in order to improve patient care management.

Rob DeConde
Combining orthogonal data sets in an effort to reconstruct biological networks (whether involving genes and expression, proteins and interaction, genetics and growth, or a combination of such components) is a cornerstone in generating predictive models in systems biology. Such networks, once created, can be used in predicting disease outcomes, identifying genetic pathways, or verifying and refining high-throughput data, among a myriad of other uses.

There are several approaches to reverse engineering networks from data, including Bayesian networks, Boolean networks, covariance structure analysis, and others. These methods all struggle in situations where the number of samples is significantly less than the number of measured variables, which is common in biological experiments. My research, currently focused in the field of covariance structure analysis, examines ways to overcome this high-dimensional obstacle algorithmically and through the incorporation of orthogonal data as a constraint on the problem (e.g. adding transcription-factor binding information to constrain network predictions based upon expression data). I test the methods via their performance on synthetics networks and predictions in biological systems.

M. Raafat El-Gewely
Genetic approach tools to protein engineering and folding with potential applications in therapeutics development. A novel method for Gene Expression profiling based on the non-randomness of the genetic code is being validated and possibly multiplexed and automated. In contrast to microarray-based methods, this method is unbiased. Currently involved in the Interferon induced transcriptional regulation with Timothy Ravasi, Merril Gersten, Kate Licon and Trey Ideker

Merril Gersten
My research has focused on using a systems approach to study neuronal molecular mechanisms underlying HIV associated dementia. By integrating rhesus brain expression data with a human protein-protein interaction network I have identified downregulation of an early response transcription factor, EGR1, as a likely contributor to the neural dysfunction seen in SIV encephalopathy.
I am recently submitted a manuscript on this research: Gersten, M., Marcondes, C., Flynn, C., Alirezaei, M., Ravasi, T., Ideker, T., and Fox, H. "An integrated systems analysis implicates EGR1 downregulation in SIVE-induced neural dysfunction".

Greg Hannum and Rohith Srivas
We are studying the use of genome-wide association data to elucidate functional relationships between protein complexes and functional groups. These results are being compared to genetic networks produced from synthetic knockouts in yeast.

Eric Jaehnig
Regulation of the transcriptional response to DNA damage by phosphorylation
I'm a postdoctoral fellow working on a joint research project between the laboratories of Dr. Richard Kolodner and Dr. Trey Ideker, using systems biology approaches to investigate the role of transcription in the cellular response to DNA damage. My project involves using mass spectroscopy to identify DNA damage-induced phosphorylation sites on transcription factors in yeast. To determine the roles of these sites in DNA damage response, I am evaluating phosphorylation site mutants for defects in DNA damage repair. Finally, I am employing a variety of approaches to identify and test candidate kinases and phosphatases that may regulate these sites. My ultimate goal is generate a phosphorylation network for the pathways regulating the transcriptional response to DNA damage.

Ryan Kelley
The general focus of my research is the combination of high throughput biological data sources (such as deletion fitness profiling and microarray expression data) with interaction data (such as yeast two-hybrid screens and genome-wide location analysis) to identify biological pathways of specific interest. Specifically, my main research focus is adaptation to oxidative stress, a process with implications for disease and aging. This work requires the generation and statistical analysis of a variety of complex data sets. As part of this work, I have developed a maximum likelihood approach to detecting the differential expression of genes from microarray expression data. In addition, I am the current maintainer of JActiveModules, a software tool which identifies active pathways from gene expression data in combination with other interaction information.

KiYoung Lee
Methods for discovering protein functions across multiple species in a genome scale based on protein networks
Rapid development of DNA sequencing technology has provided a number of fully sequenced genomes of diverse organisms. Currently, there is a high need for computation methods to automatically predict the functions of the newly sequenced genes. To fulfill the needs we are developing new methods for protein function prediction, including subcellular localization [Lee et al, NAR 2008], mainly based on protein interaction networks across multiple species. Most importantly, a protein may have different functions depend on distinct conditions such as various external stresses, disease developmental stages, and/or cell differentiation stages. These endogenous or exogenous conditions significantly influence proteins’ function. Now, we are focusing on developing computational methods for prediction of dynamic functions of human proteins under diverse disease or differentiation stages in a genome-wide scale.

Keiichiro Ono
I am developing a network visualization/analysis platform called Cytoscape (http://www.cytoscape.org). My main interest is integration of biological data over distributed database/web services and its visualization on Cytoscape.

Fan Zhang
Transcription factors network in the S.Pombe for DNA damage response
By profiling the transcription fators in S. Pombe by Chip-chip and Chip-seq with and without DNA damge agents, the cellular circuit of genes involved in the DNA damge response will be constructed.