Computational and Statistical Genomics Branch (CSGB)

Alexander Wilson, Ph.D.

Alexander Wilson
Co-Chief & Senior Investigator
Computational and Statistical Genomics Branch

Genometrics Section

B.A. Western Maryland College, 1975
Ph.D. Indiana University, 1980

phone (443) 740-2918
fax (443) 740-2165
Johns Hopkins University
333 CASSELL DR, Suite 2000

Selected Publications

Genometric Analysis Simulation Program (GASP)

Along with his background in medical genetics and biology, Dr. Wilson uses statistical, mathematical, and computer science approaches to develop new methodologies for performing statistical genetic analysis. He studies continuous or quantitative traits that are caused by both genetic and environmental factors, which are quite different from simple discrete single-gene Mendelian disorders. By analyzing the patterns of hundreds of thousands to millions of genetic markers or variants, Dr. Wilson's group identifies chromosomal regions where the genes for these traits most likely reside.

Dr. Wilson's research covers a wide range of disorders, from scoliosis (extreme curvature of the spine) to obesity and cardiovascular disease. Working with investigators at the University of Colorado, Dr. Wilson's group recently identified five regions that may be responsible for the development of scoliosis, two of which have already been confirmed by other researchers. These discoveries are significant, because scoliosis affects about one in 200 people, most often girls between 10 and 16 years of age. Although many cases of scoliosis are mild, some can be crippling.

Dr. Wilson's group has also been involved with Investigators from the Sequenced Treatment Alternatives to Relieve Depression study in a large collaborative effort aimed at finding genetic factors underlying response to a drug used in treating major depression. In this study, a genetic marker was found in the serotonin 2A receptor 5HTR2A that was associated with patients' response to citalopram, a selective serotonin reuptake inhibitor. Earlier animal studies have shown citalopram to downregulate the 5HTR2A receptor. The identification of this and other markers may lead to a personalized treatment for individuals with depression, a significant departure from the current trial-and-error approach to prescribing antidepressive medications.

Dr. Wilson also helps to develop important new methodologies to bolster statistical geneticists' toolkits. For example, he combined a traditional test of heritability with a standard analysis of variance test in a way that simplifies and significantly reduces the cost of testing for the heritability of quantitative traits. This methodology is called Regression of Offspring on Mid-Parent (ROMP). Tests of association for quantitative traits traditionally have required genotyping parents and offspring in large numbers of families, a process that can be extremely costly. However, ROMP requires investigators only to genotype the offspring; obtaining phenotypes of the parents is sufficient when using this method. In a study of high blood pressure, for example, scientists would use ROMP to genotype the offspring while only checking the parents' blood pressures. ROMP is then used to estimate the heritability of the trait and determine whether the locus being studied contains a gene that affects the trait.

Recent theoretical efforts focus on the development of linear regression methods that use multiple and/or stepwise regression in regions that are bounded by recombination hotspots (areas of increased recombination) over the entire genome. This "tiled regression" approach is being used to test for trait-marker associations in genome-wide association studies and in sequence data. It allows for the inclusion of genetic markers that are physically very close together (in linkage disequilibrium). With this approach, it becomes practical to analyze millions of markers and their significant gene x gene interaction terms.

Dr. Wilson also created a software program called GASP (Genometric Analysis Simulation Program), which enables scientists to create artificial populations or families with different mixtures of genetic and environmental influenced diseases. Because real data are often "noisy," GASP allows the creation of sample situations without extraneous factors, with one or more genes plugged in for analysis by various statistical methods. In this way, statistical geneticists can use GASP to try out new analytical approaches. Investigators at more than 70 institutions in at least 14 countries are using GASP to test new methodologies and as a teaching tool.

Top of page

Last Reviewed: December 26, 2011