A Functional Genomic Study on NCI's Anticancer Drug Screen

Abstract

Pharmacogenomics requires massive computer exploration on heterogeneous databases. The Developmental Therapeutics Program (DTP) at the National Cancer Institute (NCI) launched an unprecedented effort to build up a comprehensive drug sensitivity database about two decades ago. Sixty representative human cell-lines from seven cancer types (lung, colon, melanoma, kidney, ovary, brain, leukaemia) were selected and their responsiveness over a broad concentration range for tens of thousands of anti-cancer compounds were tested. The high volume assay has resulted in the identification of many current cancer drugs. With the project evolving and expanding over years, more cell-lines and compounds have been tested and the database is continuously updated. On a different front, the recent advance of high throughput microarray technology offers a novel molecular characterization of the cell lines. Correlating the drug sensitivity database with the gene expression database thus marks a path- breaking post-genome development in the molecular pharmacology of cancer. However, it remains puzzling that many drugs of known mechanism turn out uncorrelated with their molecular-target gene expression. We develop a system for identifying candidate genes that intervene, confound and weaken the drug- gene correlation. Our results link Methotrexate sensitivity to DNA component biosynthesis, Taxol sensitivity to genes associated with HIV-infection, and the human prion to the expression network of Alzheimer disease. The key element in our system, a novel statistical notion called "liquid association", will be discussed in detail.