Artificial Intelligence Research Laboratory Department of Computer Science Iowa State University

Algorithmic Tools for Gene Expression Analysis
   Personnel   Project Summary   Funding   Publications   Additional Information   Projects   AI Lab  

• Dr. Vasant Honavar, Principal Investigator.
• Drena Dobbs, Co-Principal Investigator.
• Pat Schnable, Co-Principal Investigator.
• Philip Haydon, Co-Principal Investigator.
• Mr. Adrian Silvescu, Ph.D. Student.

Project Summary

The central dogma of modern biology states that the functional state of an organism is determined largely by the pattern of expression of genes. Thus, the function of a cell, how well a cell performs its function, and even the determination of a cell's type is controlled by the level at which genes are expressed in the cell. Different gene expression levels can, for example, differentiate a neuron from a muscle cell or a normal cell from a cancerous cell of the same type. However, many biological processes of interest (e.g., cellular differentiation during development, aging, disease) are controlled by complex interactions over time between hundreds of genes. Furthermore, each gene is involved in multiple functions. Given the fact that thousands of genes are involved in determining the functional state of an organism, the task of assigning functions to genes, identifying underlying genetic signalling pathways, genetic control and regulatory networks is a formidable task.

The advent of microarray technology provides biologists with the ability to measure the expression levels of thousands of genes in a single experiment. Microarrays come in several flavors. In a common type of microarray, typically called a DNA microarray, several thousand DNA samples are fixed to a glass slide, each at a known position in the array. Each probe (a DNA sample) corresponds to a single gene within the organism under investigation. Messenger RNA samples are then collected from a population of cells under a given set of experimental conditions. These samples are converted into cDNA via reverse transcription and labeled with a fluorescent dye. The cDNA sample is hybridized with the microarray. The level of expression of a particular gene is roughly proportional to the amount of cDNA that hybridizes with the DNA affixed to the slide. By repeating this process under different experimental conditions, using different fluorescent dyes to distinguish between any pair of experimental conditions, it is possible to measure the expression levels of different thousands of genes under the chosen conditions. Thus, for example, the relative levels of expression of a particular gene under any pair of experimental conditions can be obtained by measuring the ratio of the amount of the two dyes present at the position of each DNA sequence on the slide using laser scanning technology.

With the increasing use of DNA microarray and related technologies for gathering gene expression data from plants and animals, there is a growing need for sophisticated computational tools for extracting biologically significant information from gene expression data, assigning functions to genes, and identifying shared signalling pathways and control circuits (e.g., genetic signalling and genetic regulatory networks). Against this background, our research is aimed at precise characterization of the information requirements of these tasks and the design and implementation of suitable algorithms for gene expression analysis. Algorthms are being developed for:

• Clustering of gene expression patterns to identify genes that are coexpressed and possibly coregulated.
• Prediction of gene function based on expression patterns.
• Discovery of statistically significant associations between expression patterns of a chosen gene and a subset of other genes.
• Inference of unknown genetic signalling networks from gene expression time series data.

The resulting computational tools for gene expression analysis will be applied to a variety of problems in computational biology including:

• Understanding the relationship between gene expression in the nervous system (e.g., in different neuron and glial cells) under various developmental, pathological, and functional states (in collaboration with Phil Haydon's laboratory).
• Identification of co-regulated genes, assignment of genes to functional families based on expression patterns, and working out underlying genetic signalling pathways in plants (especially maize and arabidopsis in collaboration with Pat Schnable's laboratory).

Funding

• Algorithmic Tools for Gene Expression Analysis. Carver Foundation 2000-2001. Vasant Honavar, Pat Schnable, Phil Haydon, and Drena Dobbs. $30,000.
• A Gene Specific DNA Chip for Exploring Molecular Evolutionary Change, Carver Foundation, (1998-1999), Gavin Naylor and Vasant Honavar. $17,200.
• IGERT: Computational Biology Training Program. National Science Foundation (1999-2004). Daniel Voytas, Susan Carpenter, Vasant Honavar, Patrick Schnable, Jonathan Wendel. $2,374,597 (plus $1,161,010 in matching funds).

Publications

• Silvescu, A. and Honavar, V. (2001). Temporal Boolean Network Models of Genetic Networks and Their Inference from Gene Expression Time Series. In: Proceedings of the Atlantic Symposium on Computational Biology, Genome Information Systems & Technology. In press.
• Honavar, V., Silvescu, A., Koul, N., Dobbs, D., Haydon, P. and Schnable, P. (2001). Algorithmic Tools for Gene Expression Analysis. To appear.

Additional Information

To appear.

© Vasant Honavar, 1999, 2000.