SCIENTIFIC RATIONALE: Large-scale evidence shows that pharmacogenetic (PGx) testing improves clinical outcomes [PMID:36739136]. However, many PGx variants and their interactions with other factors remain unknown [PMID:36707729]. Large-scale population biobanks coupled with electronic health records, including prescriptions, offer new possibilities for high-throughput PGx research [PMID:38633781]. We hypothesise that, among other traits, age and stress-related endocrine status are important mediators of drug response.
RESEARCH QUESTION: Which genetic variants and measured phenotypes associate with adverse reactions and therapeutic success?
AIMS AND OBJECTIVES: (1) To build a database of drug-response profiles from EHRs (e.g. adverse reactions, dose adjustments, therapy lengths, medication switches) for drugs acting on cardiovascular and nervous systems; (2) to build a database of glucocorticoid receptor (GR) activity/HPA axis-related phenotypes as readout of endocrine status (3) to perform genome-wide multi-trait association studies using novel methods [https://doi.org/10.1101/2025.04.29.25326633] to extract both genetic variants as well as other measured factors (e.g. age, other medication, GR/HPA phenotypes) associated with drug response phenotypes. We will include any relevant covariates (e.g., local population stratification).; (4) to build multifactor predictive models to predict drug response phenotypes from step 1; (4) to study in detail how age and HPA axis activation interact with PGx variants and how they contribute to prediction.
OUTCOMES: Our integrated approach should reveal novel markers and generate testable hypotheses for pharmacogenomics. Results will be published at 2+ international conferences and in 3+ peer-reviewed publications. All of the code generated will be released to the scientific community.