Adjunct Assistant Professor of Biostatistics
- Ph.D. Statistics (2007) UCLA
I am interested in developing statistical methods for complex disease research. Many complex diseases such as diabetes, breast cancer, chronic fatigue and obesity result from a combination of inheriting susceptibility alleles and exposure to environmental factors. One explanation for the high prevalence of many complex diseases is that risk alleles are common but each confers only a small genetic effect. The disease develops when an individual inherits several of these common risk alleles. Common small-effect alleles are difficult to study because it can be difficult to distinguish true signal from noise. While rare large-effect alleles resulting in Mendelian diseases can be tracked through pedigrees, this strategy has proven less effective for complex diseases. Instead, it may be more beneficial to study the genetics of individuals at the population level.
Preliminary study of complex disease etiology might require 1,000's to 10,000's of subjects for genome-wide association analysis. The large number of subjects, coupled with modern high-throughput genotyping (with 500k - 1 million SNP chips) and expression profiling (60k microarrays) technology can exceed the capacity of many modern genetic analysis software.
Genetic networks, which identify higher order structures within genetic data, may offer a means for a more tractable analysis. This approach identifies clusters of genes that have similar properties and then relates these clusters rather than the individual genes to clinical traits of interest. The thought is that these gene clusters represent biological pathways, and by identifying a disease related pathway, one can then target the genes that have risk alleles. My work here has mostly involved using correlation to construct gene co-expression networks and integrate them with other genetic and clinical data. Possible future work includes developing models for this algorithmic approach.
For a full list of publications, pleas see here.