Public Release:  New genetic analysis reveals principles of phenotypic expression

American Institute of Physics

College Park, MD (June 22, 2010) -- The Human Genome Project, along with numerous parallel efforts to solve the DNA sequences of hundreds of animal, plant, fungal, and microbe genomes in the last few decades, has produced enormous amounts of genetic data with which researchers are struggling to keep pace. Knowing gene sequences, after all, may not directly reveal what roles that genes play in the actual manifestation of physical traits (or phenotypes) of an organism -- including their roles in human diseases. To help navigate the new genomic landscape, researchers are developing experimental approaches and analysis tools to help prioritize and organize complex genetic information with respect to phenotypic effects.

In the journal Chaos, which is published by the American Institute of Physics (AIP), scientists at the University of Alabama at Birmingham report powerful new techniques for studying the phenotypes related to genetic differences in the budding yeast, Saccharomyces cerevisiae. The researchers took yeast cultures from an extensive library of approximately 5,000 mutated strains and subjected them to hydroxyurea -- an anti-cancer drug with known effects on the cell cycle.

Using a method called quantitative high-throughput cellular phenotyping (Q-HTCP), the researchers analyzed growth curves for tens of thousands of individual cultures, "focused on finding all of the genes that modulate the cellular effects of the drug," says study co-author John Hartman, an assistant professor of genetics. The researchers then selected the 300 "most 'hydroxyurea-interactive' genes" and further classified the genes by testing their influence on cell growth after treatment with drugs acting by different mechanisms. To integrate the results from such experiments, the researchers developed a new data mining approach called Recursive Expectation-Maximization Clustering (REMc). The approach, Hartman says, "has advantages over prior methods with respect to defining cluster number and quantifying cluster quality," which augments biological discovery.

The technique, Hartman adds, "offers a new way for trying to understand how genetic variation -- such as that related to human disease -- is alternatively buffered or expressed." Understanding phenotypic expression at a systems level, he says, would help create a new field of medicine, dubbed "phenomics."

###

The article, "Recursive Expectation-Maximization clustering (REMc): A method for identifying buffering mechanisms composed of phenomic modules" by Jingyu Guo et al will appear in Chaos: An Interdisciplinary Journal of Nonlinear Science. See: http://chaos.aip.org/

Journalists may request a free PDF of this article by contacting jbardi@aip.org.

ABOUT CHAOS

Chaos is an interdisciplinary journal of non-linear science. The journal is published quarterly by the American Institute of Physics and is devoted to increasing the understanding of nonlinear phenomena and describing the manifestations in a manner comprehensible to researchers from a broad spectrum of disciplines. Special focus issues are published periodically each year and cover topics as diverse as the complex behavior of the human heart to chaotic fluid flow problems. See: http://chaos.aip.org/

ABOUT AIP

The American Institute of Physics is a federation of 10 physical science societies representing more than 135,000 scientists, engineers, and educators and is one of the world's largest publishers of scientific information in the physical sciences. Offering partnership solutions for scientific societies and for similar organizations in science and engineering, AIP is a leader in the field of electronic publishing of scholarly journals. AIP publishes 12 journals (some of which are the most highly cited in their respective fields), two magazines, including its flagship publication Physics Today; and the AIP Conference Proceedings series. Its online publishing platform Scitation hosts nearly two million articles from more than 185 scholarly journals and other publications of 28 learned society publishers.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.