Combining multiple knowledge sources: A case study of drug induced liver injury

Abstract

Many classes of drugs, their interaction pathways and gene targets are known to play a role in drug induced liver injury (DILI). Pharmacogenomics research to understand the impact of genetic variation on how patients respond to drugs may help explain some of the variability observed in the occurrence of adverse drug reactions (ADR) such as DILI. The goal of this project is to combine rich genotype and phenotype data to better understand these scenarios. We consider similarities between drugs, similarities between drug targets, drug-pathway-gene interactions, etc. Links to the patients will include patient drug usage, ADR, disease outcomes, etc. We will develop appropriate protocols to create these rich datasets and methods to identify patterns in graphs for explanation and prediction.