With the increasing burden of multi-morbidity and polypharmacy, there is an unprecedented need to develop efficacious and safe therapies to prevent and treat disease. However, drug development has a notoriously high failure rate, with less than 10% of drugs entering clinical study eventually being approved for use in patients. This low success rate may be at least partly attributed to most of the pre-clinical evidence for drug effects coming from animal studies and traditional epidemiological associations, which are respectively limited in their translatability to humans and their ability to draw causal inferences.

Human genetics offer the opportunity to dramatically improve the efficiency of drug development. The majority of drug targets are proteins, which are in turn coded for by genes. Naturally occurring variation in the genes coding for these proteins can mimic the effects of drugs targeting them. By studying associations of such genetic variation in large-scale population data, it is possible to infer the effect of therapeutically targeting these proteins. Judicious use of human genetic data in this ‘drug target Mendelian randomization’ paradigm has the potential to inform on various critical aspects of drug development, including on-target efficacy, safety, repurposing, biomarker selection, effect heterogeneity, and interactions.

Mendelian randomization can complement more traditional pre-clinical research efforts because it uses genetic data to infer the causal effects of drug target perturbation in humans. Drug targets with human genetic evidence have been shown to be more than twice as likely to make it through clinical development. Given that the average new drug requires more than 10 years and 1 billion US dollars to obtain regulatory approval, such insights have the potential to tremendously improve the efficiency of generating effective treatments to prevent and treat disease.