Fred A. Wright
School of Public Health
Area of interest
Fred Wright, Ph.D., is a statistical geneticist in the UNC Department of Biostatistics, with wide-ranging interests in genomics and bioinformatics. He has collaborated with cancer researchers for his entire career, and was a member of the Cancer Centers at UC San Diego and Ohio State University in previous positions.
Wright has worked for a number of years on the development of statistical methods in linkage mapping, an important component of modern genetic epidemiology. The objective of linkage mapping is to use the segregation of genetic markers in a pedigree to obtain an approximate chromosomal location for a gene of interest. His current interests in this area include further development of methods for mapping quantitative traits, and other outstanding problems include the simultaneous identification and mapping of multiple genes that contribute to a single trait. In addition, he is working on new methods for the identification of haplotype-disease association. All of these approaches are relevant to the search for genes involved in cancer susceptibility, which has increasingly focused on the identification of low-penetrance genes and complex disease models.
In addition, Wright has been pursuing bioinformatics problems enabled by the availability of the draft human genome sequence. Together with colleagues at the Ohio State University, he continues working on whole-genome assembly techniques and the use of assembled Unigene EST clusters to identify likely disease gene candidates (including cancer).
Much of Wright's recent research effort has been devoted to the analysis of gene expression microarrays. He has developed new methods for estimating gene expression from Affymetrix oligonucleotide arrays, and has analyzed several cancer data sets to identify genes differentially expressed among histological subtypes. This work will be a primary focus for the next several years, and Wright has postdoctoral personnel and Ph.D. students actively pursuing improved methods for analyzing array data.
Finally, Wright is actively working on methods to formally localize putative tumor suppressor genes using loss-of-heterozygosity (LOH) data. By using likelihood-based approaches, his group can rigorously analyze LOH data and synthesize the results across multiple studies. He has completed such a pooled analysis of over 150 studies in breast cancer, and this will be followed by similar analyses of additional tumor types. This work is also informing the development of a similar statistical framework to analyze data arising from newer technologies such as array-based comparative genomic hybridization.
Awards and Honors
In 2001, Wright and Dr. Bo Yuan of Ohio State University published two papers that have had a major impact on the genomics community. By computationally assembling expressed sequence tags from the UniGene collection (Zhou et al., 2001), the group produced one of the first human gene maps. A follow-up paper (Wright et al., 2001) applied this EST assembly and many other resources to the emerging consensus human DNA sequence, resulting in one of the first draft annotations of the genome. Wright et al. (2001) is listed as one of the all-time top 10 most-accessed articles in PubMed Central. All of this work has direct cancer relevance in the identification of candidate genes involved in the etiology of cancer. Recently Wright's group has completed an analysis of loss of heterozygosity in breast cancer (Miller et al., in press), providing much-needed clarification of the degree of evidence for preferential allelic loss in defined genomic regions.