School of Informatics and Computing Menu

Pankaj Agarwal, Ph.D.: Informatic Discovery of Disease Indications for Drugs



Proving that a drug that has seen human patients also works in another indication adds clear value for patients. The question for informatics is can these predictions be made systematically with precision. We will discuss two recent methods that add to this debate. The first one uses genetics associations (from Genome Wide Association Studies – GWAS) to find new indications for drugs some of which are being validated experimentally. The second one frames the side effects of drugs as pharmacological effects, relies on the observation that many drugs that treat a disease have similar side effects, and then predicts other drugs with similar side effects as potential treatments for the disease. In addition, we will discuss other techniques that use text mining, compound-based methods such as connectivity map, electronic health records, and other public data. We hope that the field of drug repositioning will continue to be a showcase for informatics techniques in medicine.


Pankaj Agarwal obtained a B. Tech. in Computer Science & Engineering from IIT, Delhi. His Ph.D. is in Computer Science from the Courant Institute, New York University. His thesis was on “Cell-based Computer Models in Developmental Biology”. Pankaj did a postdoc in Bioinformatics at the Washington University, St. Louis, MO. He was one of the founding members and directors of the International Society for Computational Biology (ISCB). Since 1996, he has been at GlaxoSmithKline, and has led on a number of projects to mine biological data in to demonstrate informatics value for the pharmaceutical industry. He leads the Systematic Drug Repositioning group at GSK and his research is focused on computational methods for discovering the right disease indication for drugs.