Drug-induced liver injury (DILI) is a serious medical condition with important implications in public health and drug development. Genetic studies of DILI aim to uncover genes and variants that predispose risk, which could shed light on the biological mechanism of liver injuries and help developing genetic markers for personalized treatment. Here I describe our effect to map risk variants and genes of augmentin-induced liver injury. The cases were collected at several clinical networks and compiled by the International Serious Adverse Events Consortium (SAEC). Firstly we looked at common polymorphisms in human populations by a genome-wide association study (GWAS). By this approach we found significantly associated common variants in the Major Histocompatibility Complex (MHC) region with strong effects, confirming the importance of immune system in DILI. However, although about 80% of cases are carriers of the major implicated variants, less than 1% of carriers in the general population are actually at risk. That is similar to the missing heritability problem we face in genetic studies of complex diseases. One of promising directions is to investigate the contribution of rare variants that were under the radar of GWAS. Therefore we are carrying out exome sequencing on the same set of cases. In exome sequencing, we directly interrogate all genomic variants in coding regions and search for genes that increase disease risk by group-wise association tests. I will talk about the details of analyzing exome data and the preliminary results from the case-control association. Together, these two studies interrogated genetic variants from the entire frequency spectrum, found genetic variants and genes that increase the risk of drug-induced liver injury, and provide potential basis of genetic tests for personalized treatment.