Human Genome Variation: Disease, Drug Response, and Clinical Phenotypes - Session Introduction

With the completion of a rough draft of the human genome sequence in sight, researchers are shifting to leverage this new information in the elucidation of the genetic basis of disease susceptibility and drug response. Massive genotyping and gene expression profiling studies are being planned and carried out by both academic/public institutions and industry. Researchers from different disciplines are all interested in the mining of the data coming from those studies; human geneticists, population geneticists, molecular biologists, computational biologists and even clinical practitioners. These communities have different immediate goals, but at the end of the day what is sought is analogous: the connection between variation in a group of genes or in their expression and observed phenotypes. There is an imminent need to link information across the huge data sets these groups are producing independently. However, there are tremendous challenges in the integration of polymorphism and gene expression databases and their clinical phenotypic annotation This is the third session devoted to the computational challenges of human genome variation studies held at the Pacific Symposium on Biocomputing 1,2. The focus of the session has been the presentation and discussion of new research that promises to facilitate the elucidation of the connections between genotypes and phenotypes using the data generated by high-throughput technologies. Nine accepted manuscripts comprise this year's original work presented at the conference. A major incentive for collecting genetic variation data is to use this information to identify genomic regions that influence disease susceptibility or drug response. In this volume, Zhang et al. outline a new approach to identify clinically relevant genes that produce quantitative phenotypes. Although similar methods have been developed to measure the strength of association between haplotypes and binary (case-control) data, Zhang et al.'s method is particularly valuable because many