Classification of non-coding variants with high pathogenic impact

Abstract

Whole genome sequencing is increasingly used to diagnose medical conditions of genetic origin. While both coding and non-coding DNA variants contribute to a wide range of diseases, most patients who receive a WGS-based diagnosis today harbour a protein-coding mutation. Functional interpretation and prioritization of non-coding variants represents a persistent challenge, and disease-causing non-coding variants remain largely unidentified. Depending on the disease, WGS fails to identify a candidate variant in 20–80% of patients, severely limiting the usefulness of sequencing for personalised medicine. Here we present FINSURF, a machine-learning approach to predict the functional impact of non-coding variants in regulatory regions. FINSURF outperforms state-of-the-art methods, owing in particular to optimized control variants selection during training. In addition to ranking candidate variants, FINSURF breaks down the score for each variant into contributions from individual annotations, facilitating the evaluation of their functional relevance. We applied FINSURF to a diverse set of 30 diseases with described causative non-coding mutations, and correctly identified the disease-causative non-coding variant within the ten top hits in 22 cases. FINSURF is implemented as an online server to as well as custom browser tracks, and provides a quick and efficient solution to prioritize candidate non-coding variants in realistic clinical settings.

PSAP-genomic-regions: a method leveraging population data to prioritize coding and non-coding variants in whole genome sequencing for rare disease diagnosis

“…The method most comparable to the strategy followed by PSAP-genomic-regions is the recently-developed machine-learning algorithm FINSURF (42). FINSURF aims to predict the functional impact of non-coding variants in regulatory regions and has been applied to known pathogenic variants inserted in WGS data like we did.…”

Identification of ancestral gnathostome Gli3 enhancers with activity in mammals

“…Over the last few years, it has become increasingly clear that the majority of disease‐associated mutations reside within noncoding regulatory enhancer regions (Claringbould et al, 2021; Kvon et al, 2020; Moyon et al, 2022). For instance, based on GWAS data, a SNP rs17640804 (C > T) in the GLI3 noncoding region (7q13, intron 3) has been associated with human face morphology, more specifically, with determining the nose wing breadth (Adhikari et al, 2016).…”

A rare non-coding enhancer variant inSCN5Acontributes to the high prevalence of Brugada syndrome in Thailand

“…To assess for linkage disequilibrium between variants, r 2 values were calculated for each variant pair in this region (with a combined minimum case and control allele count of 3). Variants of interest were further annotated with CADD and FINSURF scores 13 , frequency in gnomADv3 East Asian genome sequences (n=2,604) and conservation of affected base across species using Ensembl v109 (Feb 2023). The frequency of the putative causal variant was also assessed in other genomic databases with individuals of Asian ancestry – the Genomics Thailand database of 10,043 individuals of Thai ancestry, the Singaporean SG10K_Health v5.3 database of individuals of Chinese, Indian and Malay ancestry 33 , the jMORP/54KJPN database of individuals of Japanese ancestry 34 and the KOVA database of individuals of Korean ancestry 35 .…”

“…The RE5 variant alters a nucleotide fully conserved across 35 mammalian species with available orthologous sequence data, in contrast to the other three variants (Fig. S1) and has higher computational deleteriousness scores from the CADD and FINSURF 13 algorithms (Table 2). The RE5 variant is also absent in the gnomADv3 database (n=76,165 genomes including n=2,604 of East Asian (EAS) ancestry), whereas the SCN10A -short promoter variant has a gnomAD-EAS frequency of 0.001, indicating it also exists on haplotypes distinct from the RE5 variant.…”