Faculty and Student Teams Program

Project Descriptions

Cellular and Molecular Interactions among Genes in
Cerebellar Development over Time and Space

Requesting applications from science or engineering faculty members at institutions serving students underrepresented in science, engineering, mathematics and technology to work on the following projects at Oak Ridge National Laboratory (ORNL).

Project Description

This project aims at determining the cellular and molecular interactions among genes in cerebellar development over time and space. Logical networks allow the representation of a gene regulatory network via a causal computational model. In a logical network, a logical function associated with each gene as a node describes its behavior dictated by regulator genes. In inference of a logical network, continuous gene expression levels will be quantized to binary or tertiary with an optimal quantization technique that maximizes either a joint or a marginal likelihood. Potential regulators of each gene will be selected using the transcriptome quantitative trait locus (QTL) mapping as a positional filter. The optimal logics at each gene node in the network will be searched so that they best explain the overall interactions. When only steady-state gene expression data are available, the strategy is to search for optimal logics under the constraint that they will bring attractors of the sought-after logical network closer to the quantized observed gene expression levels.

A logical network can be considered a transition model from quantitative probabilistic networks such as Bayesian networks to dynamical system models such as difference equations. As with a Bayesian network, nodes in a logical network can represent gene expression levels as well as experimental conditions. The functions at each node in a logical network provide causal descriptions of interactions among genes, while the conditional probability functions at each node in a Bayesian network do not. Cyclic controls are common in nature. The mathematical formulation of Bayesian networks cannot elucidate cycles directly, while cycles are explicit in a logical network. Logical networks are dynamic but they can be approximated with steady-state gene expression data sets. Therefore, a logical network can serve as an intermediate model to delineate an anticipated dynamics of gene interactions when a more complex dynamical model based on time series is infeasible due to either experimental or computational considerations. The application of logical networks to the genetic analysis of gene expression data will facilitate extension to multi-condition data sets including time series and experimental treatments including exposures to environmental agents.

Laboratory Contact: Dr. Elissa Chesler, cheslerej@ornl.gov

Applicants Responsibilities and Relationship to Project

Applicants will receive support under the Department of Energy Faculty Student Team Research Program (FaST) to work collaboratively with the project research team at the Laboratory for up to 10 weeks during the summer of 2008. The exact appointment period in the time frame of June to August will be scheduled by mutual agreement between the host divisions at Oak Ridge National Laboratory and the successful applicant. Faculty will be expected to identify students from their campuses to participate in the FaST program. The faculty member will provide some mentorship to students during the summer research activities. The faculty and students must participate as a group and serve their appointments concurrently. It is expected that the faculty member and the students become an integral part of the research team working on this project and that opportunities for continued collaboration may be identified.

Qualifications of Ideal Candidates

The ideal faculty candidate for this position has background and expertise to carry out self-directed research on network analysis of large-scale systems genetics data sets originating from ORNL’s mouse reference populations. The candidate should have a background in computer science with experience in systems biology. Specifically, an understanding of the biology of gene regulatory network is required. Quantitative analysis skills such as probabilistic network models and dynamical system modeling are essential. Aspects of the research that the candidate will investigate and support include using probabilistic and dynamical models to quantify the interactions among genes from their expression data. The candidate should have programming experience with C/C++, python, and a statistical computing language such as R. Further, they should be familiar with tools for QTL analysis including GeneNetwork.org and R/QTL.

Support and Financial Commitments

The successful candidate will receive a stipend based on the academic salary, travel expenses to and from the Laboratory, and a housing allowance. Students recommended by the faculty member for participation in the program will receive a stipend of $400/week for each week at the Laboratory, plus a housing allowance, and reimbursement for transportation expenses to and from the ORNL. Funds are provided for this program from the US Department of Energy, Office of Science in partnership with the National Science Foundation, from ORNL, and from other sources.

See Financial Information.