Augmenting Pharmacogenetics with Multi-Omics Data and Techniques to Predict Adverse Drug Reactions to NSAIDs

摘要 非类固醇抗炎药(NSAIDs)是一类广泛用于治疗疼痛、发热、 和炎症。非甾体抗炎药的使用与轻微的和危及生命的药物不良反应(ADRs)有关。 包括胃肠道出血和急性冠脉综合征。考虑到这些结果的严重性和 在庞大的患者群体中，非常需要根据非类固醇抗炎药来预测个体的ADR风险。药物遗传学 是研究遗传学如何影响药物反应的学科。临床药物遗传学的实施 该联盟已经发布了针对临床医生的指南，以在存在CYP2C9丢失的情况下修改非类固醇抗炎药的治疗方法。 非功能变异体，导致非类固醇抗炎药清除减少和不良反应风险增加。然而， 仅用CYP2C9一项就能解释ADR风险的相对较小比例，目前使用 临床协变量，如年龄、性别、伴随药物和合并症。我们建议更好地理解 非甾体抗炎药不良反应的可遗传风险研究 并利用它来开发多基因风险评分(PRS)。此外，为了改善交叉血统 ，我们将根据推定的转录本开发转录风险评分(TRS)。 并将其与我们的PRS集成在一起。然后我们将建立一个结合药物遗传学的多模式模型， 基因组学和临床变量来预测ADR风险。成功完成这两个目标将防止 无数非甾体抗炎药引起的不良反应，并提高我们对潜在的不良反应风险因素的理解。 除此之外，我们的工作将成为未来应用多组学来增强药物遗传学的模式， 让我们更接近“在正确的时间，为正确的患者，使用正确的药物。”

Abstract Non-steroidal anti-inflammatory drugs (NSAIDs) are a class of widely used drugs for treatment of pain, fever, and inflammation. NSAID use has been linked to both mild and life-threatening adverse drug reactions (ADRs) including gastrointestinal bleeding and acute coronary syndrome. Given the severity of these outcomes and the large patient pool, there is a great need to predict individual risk of ADR from NSAIDs. Pharmacogenetics is the study of how genetics influence drug response. The Clinical Pharmacogenetics Implementation Consortium has published guidelines for clinicians to modify NSAID treatment in the presence of CYP2C9 loss- of-function variants, which result in reduced clearance of NSAIDs and increased risk of ADRs. However, CYP2C9 alone explains a relatively small proportion of risk of ADR, which is currently better predicted using clinical covariates such as age, sex, concomitant drugs, and comorbidities. We propose to better understand the heritable risk of NSAID ADR by performing a genome-wide association study for NSAID ADR in a diverse population and using it to develop a polygenic risk score (PRS). Furthermore, to improve cross-ancestry performance of our PRS, we will develop a transcriptomic risk score (TRS) based on imputed transcriptomes and integrate it with our PRS. We will then build a multi-modal model that combines pharmacogenetics, genomics, and clinical variables to predict ADR risk. Successful completion of both aims would prevent countless NSAID-induced ADRs and improve our understanding of the risk factors underlying ADR risk. Beyond that, our work will serve as a model for future application of multi-omics to augment pharmacogenetics, bringing us closer to “the right drug, for the right patient, at the right time.”