Genetic drug target validation using Mendelian randomisation

Abstract

Mendelian randomisation (MR) analysis is an important tool to elucidate the causal relevance of environmental and biological risk factors for disease. However, causal inference is undermined if genetic variants used to instrument a risk factor also influence alternative disease-pathways (horizontal pleiotropy). Here we report how the 'no horizontal pleiotropy assumption' is strengthened when proteins are the risk factors of interest. Proteins are typically the proximal effectors of biological processes encoded in the genome. Moreover, proteins are the targets of most medicines, so MR studies of drug targets are becoming a fundamental tool in drug development. To enable such studies, we introduce a mathematical framework that contrasts MR analysis of proteins with that of risk factors located more distally in the causal chain from gene to disease. We illustrate key model decisions and introduce an analytical framework for maximising power and evaluating the robustness of analyses.

Finding drug targeting mechanisms with genetic evidence for Parkinson’s disease

“…It is therefore not possible to perform the additional quality control discussed below, which illustrates that results based on one or two SNPs should be interpreted with caution. It is noteworthy that previous eQTL-based MR studies have reported heterogeneity between tissues, both in terms of the magnitude and direction of effect (Schmidt et al 2020) We consider the MR result more robust if several meta-analysis methods yield a similar result, such as the maximum likelihood and MR-Egger methods (Haycock et al 2016;Burgess et al 2019;Slob and Burgess 2020). This is only possible if > 2 SNPs are available per gene, and we found that all eight genes with > 2 SNPs reached at least nominal significance using the maximum likelihood method (unadjusted < 0.05).…”

“…For example, an eQTL associated with reduced expression of HMGCR mimics exposure to an HMGCR-inhibitor, such as a statin. Most clinically-used drugs target proteins, and genetic variants associated with protein levels in a clinically relevant tissue may be ideal to model drug target effects with MR (Schmidt et al 2020). However, even with high throughput protein assays, the spectrum of reliable, wellpowered GWAS data on protein targets is limited.…”

“…Nevertheless, most clinically-used drugs target proteins, not gene expression. As such, genetic variants associated with protein levels, called protein quantitative trait loci (pQTLs), may model drug target effects more accurately (Schmidt et al 2020). We therefore sought to validate our 23 proposed drug targets for PD using pQTL data.…”

“…So-called "druggable" genes encode proteins that can be targeted by medications. More precisely, these are proteins which have been targeted by a pharmacological agent or are considered possible to target with a small molecule or monoclonal antibody (Finan et al 2017;Schmidt et al 2020). While genome-wide association studies (GWAS) have effectively identified single nucleotide polymorphisms (SNPs) linked to PD risk and progression (Nalls et al 2019;Blauwendraat et al 2019;Iwaki et al 2019), the GWAS design cannot reliably pinpoint causal genes nor directly inform drug development.…”

“…Mendelian randomization (MR) is a genetic technique that can predict the efficacy of a drug by mimicking a randomized controlled trial (Katan 1986;Smith and Ebrahim 2003;Hingorani et al 2005;Holmes et al 2017). MR can use genetic variants, usually SNPs, associated with expression levels of a druggable gene to mimic lifelong exposure to a medication targeting the encoded protein (Schmidt et al 2020;Zhu et al 2016). The association between the same genetic variants and a disease (the outcome) can then be extracted from a GWAS (Figure 1a).…”