Quantifying negative selection in human 3’ UTRs uncovers constrained targets of RNA-binding proteins

Abstract

Many non-coding variants associated with phenotypes occur in 3’ untranslated regions (3’ UTRs) and may affect interactions with RNA-binding proteins (RBPs) to regulate post-transcriptional gene expression. However, identifying functional 3’ UTR variants has proven difficult. We used allele frequencies from the Genome Aggregation Database (gnomAD) to identify classes of 3’ UTR variants under strong negative selection in humans. We developed intergenic mutability-adjusted proportion singleton (iMAPS), a generalized measure related to MAPS, to quantify negative selection in non-coding regions. This approach, in conjunction within vitroandin vivobinding data, identifies precise RBP binding sites, miRNA target sites, and polyadenylation signals (PASs) under strong selection. For each class of sites, we identified thousands of gnomAD variants under selection comparable to missense coding variants, and found that sites in core 3’ UTR regions upstream of the most-used PAS are under strongest selection. Together, this work improves our understanding of selection on human genes and validates approaches for interpreting genetic variants in human 3’ UTRs.

Improving estimates of negative selection in human genome using CAPS

“…Even though, as mentioned earlier, it has been shown that intergenic variants may be used as an alternative class of variants for the estimation of the baseline level of rare variation (Findlay et al, 2022), we had insufficient evidence that intergenic variants could be used interchangeably with synonymous variants for the purpose of model calibration without any significant effect on the results.…”

“…Overall, we believe that our proposed context-based approach is superior compared with using mutability as a proxy for the estimation of the expected level of rare variation, as mutation rates modelling proves a difficult task (Findlay et al, 2022). Even though it is possible to reduce bias of MAPS towards transversion variants by modifying the fit of the model, we argue that this approach can be seen as ad hoc and not future-proof, considering that every new release of gnomAD would require refitting the model.…”

“…MAPS scores are calculated as the scaled excess or deficit of singletons (variants with allele count of 1), where the expected number is derived based on context sequence from a singletons-by-mutability model calibrated on a relatively neutral class of variants. Even though it has been shown that intergenic variants can be used for calibration (Findlay et al, 2022), synonymous variants have traditionally been preferred for this purpose. Importantly, given then the complexity of mutability estimation, there is no straightforward way of modelling mutation rates; however, attempts have been made to incorporate additional mutability-related annotations in the singletons-by-mutability model (Findlay et al, 2022).…”

“…Even though it has been shown that intergenic variants can be used for calibration (Findlay et al, 2022), synonymous variants have traditionally been preferred for this purpose. Importantly, given then the complexity of mutability estimation, there is no straightforward way of modelling mutation rates; however, attempts have been made to incorporate additional mutability-related annotations in the singletons-by-mutability model (Findlay et al, 2022).…”

Systematic identification of disease-causing promoter and untranslated region variants in 8,040 undiagnosed individuals with rare disease

“…RNA binding protein predictions were generated using the methods detailed in Findlay et al ( 38 ) for all possible variants within motifs that are proximal to ENCODE eCLIP sites, that are also high affinity sites as predicted by RBPamp(39). These were intersected with our variants and filtered to retain only those with a reference affinity of ≥ 0.1 and with an impact of ‘loss of binding’ predicted by the RBP binding affinity model (defined as alternative allele affinity / reference allele affinity < ⅓).…”