The Jackson Laboratory

A personalized and predictive approach to health and disease will depend on understanding how genes and environmental factors combine to generate complex cellular behaviors. Our laboratory is developing novel computational methods to map complex genetic architecture and infer models that predict the outcomes of genetic and environmental variation. Our work involves deriving network models of interacting genes, integrating disparate phenotypic and molecular data types, critically evaluating models with experimental tests, and understanding how biological information is encoded in genetic networks and genomic data.

Modeling pleiotropy and epistasis

The observations that a gene can affect many traits (pleiotropy) and that a trait can be affected by multiple interacting genes (epistasis) imply that cellular behavior is often the result of a genetic network involving multiple biological processes. We aim to exploit patterns of pleiotropy and epistasis to infer models of how genes interact to influence multiple phenotypic measures. To this end, we are devising novel mathematical and statistical methods using baker's yeast and laboratory mice as platforms for development. By addressing the complexities of pleiotropy and epistasis we hope to improve the predictive power of network models and better reveal underlying biological mechanisms.

Integrating diverse data types

The advent of high-throughput data collection and significant advances in data curation now provide researchers with a wealth of molecular and physiological data. However, devising methods that optimize the use of these diverse data types is an ongoing challenge. We are taking an approach based in genetic modeling, in which candidate genes and interactions are identified and then used as a guide to integrating additional data on molecular interactions that may be relevant in a specific experimental system. Ideally, this translates complex genetic observations into specific molecular hypotheses of biological processes.

Predicting outcomes of novel genetic variation

The ultimate goal of our group's research is to develop methods to infer models with high predictive power. However, the translation of a genetic model to a specific prediction that can be tested in the lab is often a difficult process. We are working to devise our models in a way that makes them amenable to rigorous testing in the laboratory. This critical evaluation will identify model failures and indicate which aspects of the models need to be addressed with additional experimentation and analysis. Rounds of modeling and experimentation can be iterated until convergence on a successful model. The resulting network models will enable the testing of genetically complex hypotheses, prioritization of candidate genes for targeted intervention and the personalization of prognoses and therapies.

Quantifying information in genetic networks

The study of molecular epistasis has been used for decades in mapping pathways of linear information flow from gene to gene. However, the genetic complexity inherent in many biological systems can confound this strategy when the system is viewed on a genomic scale. Instead of mapping linear pathways, large-scale networks of genetic interactions tend to feature tangled modules of genes that function together to carry out cellular processes. Furthermore, given the prevalence and diversity of genetic interactions, it is often unclear how to optimally define the rules of genetic interaction that form the links in these networks. We are developing methods based in information theory to measure the information content of networks. This quantitative measure of complexity can serve as scoring function to find the most informative network from a given genetic data set. From this work we hope to develop both practical tools for genetic analysis and fundamental insights into how networks encode information.

Principal Investigator: Greg Carter, Ph.D.

Predoctoral Associate: Yang Zhang, B.S.

Research Administrative Assistant: Tonnya Norwood, B.S.

Carter GW, Rush CG, Uygun F, Sakhanenko NA, Galas DJ, and Galitski T. 2010. A systems-biology approach to modular genetic complexity. Chaos 20(2):026102. PMC2909309

2009

Carter GW, Galas DJ, Galitski, T. 2009. Maximal Extraction of Biological Information from Genetic Interaction Data. PLoS Computational Biology, 5(4):e1000347. PMC2659753

Carter GW, Dudley, AM. 2009. Systems genetics of complex traits. Robert, ed., Encyclopedia of Complexity and Systems Science, Springer, New York.

Killcoyne S, Carter GW, Smith J, Boyle J. 2009. Cytoscape: A Community-Based Framework for Network Modeling. Methods Mol Biology 563: 219-239.

2007

Carter GW, Prinz S, Neou C, Shelby JP, Marzolf B, Thorsson V, Galitski T. 2007. Prediction of phenotype and genomic expression for combinations of mutations. Molecular Systems Biology 3:96. PMC1847951

Selinummi J, Niemistö A, Saleem R, Carter GW, Aitchison J, Yli-Harja O, Shmulevich I, Boyle J. A case study on 3-D reconstruction and shape description of perioxisomes in yeast. Proceedings of the 2007 IEEE International Conference on Signal Processing and Communication (ICSPC 2007), 672-675.

Carter GW, Thorsson V, Galitski T. 2007. Network Modeling of Molecular and Genetic Interactions. Conn PM, ed., Source Book of Models for Biomedical Research, Chapter 9, Humana Press.

2006

Carter GW, Rupp S, Fink GR, Galitski T. 2006. Disentangling information flow in the Ras-cAMP signaling network. Genome Research 16: 520-526. PMC1457029

2005

Carter GW. 2005. Inferring Network Interactions within a Cell. Briefings in Bioinformatics 6(4): 380-389.

Drees BL, Thorsson V, Carter GW, Rives AW, Raymond M, Avila-Campillo I, Shannon P, Galitski T. 2005. Derivation of genetic interaction networks from quantitative phenotype data. Genome Biology 6: R38. PMC1088966