Human Genetic Evidence Enriched for Side Effects of Approved Drugs
Eric Vallabh Minikel, Matthew R. Nelson.
Abstract
Safety failures are an important factor in low drug development success rates. Human genetic evidence can select drug targets causal in disease and enrich for successful programs. Here, we sought to determine whether human genetic evidence can also enrich for labeled side effects (SEs) of approved drugs. We combined the SIDER database of SEs with human genetic evidence from genome-wide association studies, Mendelian disease, and somatic mutations. SEs were 2.0 times more likely to occur for drugs whose target possessed human genetic evidence for a trait similar to the SE.
Introduction
Safety issues are a major contributor to drug candidate failure, with clinical safety findings accounting for 25% of drug program terminations in Phase I-II. The causal evidence of human genetics between drug targets and phenotypic outcomes can provide insights into potential on-target safety liabilities of drug candidates before development has even begun. There are many anecdotes of adverse events predicted by genetics because they are similar to traits genetically associated with a drug target.
Materials and Methods:
Side effect (SE) data were obtained from the SIDER database (v4.1), which captures SEs from product labels and package inserts for approved drugs up through 2015. Citeline Pharmaprojects and DrugBank were parsed as described and SIDER drug names were mapped to Pharmaprojects indications using text matches to Pharmaprojects drug name synonyms, or, by mapping first to DrugBank using either ATC codes or name matches to obtain CAS numbers, and then looking up CAS numbers in Pharmaprojects; the proportion of drugs mapped by various approaches is provided in Table B in S1 Table. Pharmaprojects matches were used to obtain human gene targets and MeSH terms for approved indications as described.
Discussion
Despite encouraging anecdotes and the widespread use of human genetic data to attempt to predict safety risks, only limited systematic evidence has been reported to support the hypothesis that human genetic evidence can predict drug SEs. Our study provides evidence that SEs are roughly twice as likely to occur for a given drug if that drug’s target has a human genetic association to a trait very similar to the SE.
Citation: Minikel EV, Nelson MR (2025) Human genetic evidence enriched for side effects of approved drugs. PLoS Genet 21(3): e1011638. https://doi.org/10.1371/journal.pgen.1011638
Editor: Lynn Petukhova,, Columbia University, UNITED STATES OF AMERICA
Received: June 17, 2024; Accepted: February 26, 2025; Published: March 31, 2025.
Copyright: © 2025 Minikel, Nelson.. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: An analytical dataset and source code are available at https://github.com/ericminikel/genetics_side_effects/ and are sufficient to reproduce all figures and statistics herein.
Funding: This work was supported by Deerfield Management Company, L.P. and Genscience, LLC in the form of salaries to MRN and EVM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: I have read the journal's policy and the authors of this manuscript have the following competing interests:MRN is an employee of Genscience and Deerfield. EVM is a consultant to Deerfield Management Company, L.P. and Genscience, LLC. EVM also acknowledges research support, unrelated to the current work, from Ionis, Gate, Sangamo, and Eli Lilly, and consulting fees, unrelated to the current work, from Eli Lilly and Alnylam.
