Principal Investigator: Dr Janie Chiang
Cipherome, Inc., San Jose, USATags: 52031, adverse drug reactions, Ancestry, epidemiology, genetics/genotyping, medication prescribing, Pharmacogenomics
The goal of personalized medicine is the right drug, with the right dose, at the right time, for the right person. Pharmacogenomics may help us better understand how one’s genes process and metabolize drugs, and predict one’s responses to drugs. By utilizing the UK Biobank data, the results of our study may eventually lead to methodological improvements in recommendations of drugs and drug doses for an individual, including automatically incorporating one’s ancestry information. Our methods may also provide a way for pharmaceutical companies to improve drug development.
Despite the pharmacogenomic evaluation of many drugs and the knowledge we have of genes that are involved in drug absorption, metabolism, distribution, and excretion, there are still many important genetic biomarkers to be discovered. We do not know how all genetic variation contributes to adverse drug events. Thus, the person-to-person variability in drug response is a major challenge for current clinical practice, drug development, and drug regulation. A drug with proven clinical efficacy in some patients may fails to work in others and even cause serious side effects, including death.
We have developed an approach that provides a polygenic score for genes based on potential functional impact of rare genetic variation. We have evidence that this method may identify and predict those at risk of adverse drug reactions (ADRs). We propose to use the UK Biobank biorepository and our algorithm to evaluate if we see an enrichment for adverse drug reactions in individuals with high gene-scores. We will also compare how well our method works compared to common frequency genetic variants known to significantly impact ADR risk. Because there are a variety of genetic variants that contribute to ADR risk, we also propose to use UK Biobank data to improve our algorithm to account for the rare and common frequency genetic variants. This should improve our algorithm, leading to a more robust way for identifying individuals at risk of an ADR. If effective, this tool could assist clinicians in prescribing the right dose at the right time for the right person, and thus preventing serious side effects.
In addition, different ancestries have different frequencies of genetic variation that may contribute to ADRs. In our analyses we will also evaluate how our algorithm performs when evaluating the relationship between risk score and ADRs for specific genetic ancestries.