A comprehensive WGS-based pipeline for the identification of new candidate genes in inherited retinal dystrophies

Abstract

To enhance the use of Whole Genome Sequencing (WGS) in clinical practice, it is still necessary to standardize data analysis pipelines. Herein, we aimed to define a WGS-based algorithm for the accurate interpretation of variants in inherited retinal dystrophies (IRD). This study comprised 429 phenotyped individuals divided into three cohorts. A comparison of 14 pathogenicity predictors, and the re-definition of its cutoffs, were performed using panel-sequencing curated data from 209 genetically diagnosed individuals with IRD (training cohort). The optimal tool combinations, previously validated in 50 additional IRD individuals, were also tested in patients with hereditary cancer (n = 109), and with neurological diseases (n = 47) to evaluate the translational value of this approach (validation cohort). Then, our workflow was applied for the WGS-data analysis of 14 individuals from genetically undiagnosed IRD families (discovery cohort). The statistical analysis showed that the optimal filtering combination included CADDv1.6, MAPP, Grantham, and SIFT tools. Our pipeline allowed the identification of one homozygous variant in the candidate gene CFAP20 (c.337 C > T; p.Arg113Trp), a conserved ciliary gene, which was abundantly expressed in human retina and was located in the photoreceptors layer. Although further studies are needed, we propose CFAP20 as a candidate gene for autosomal recessive retinitis pigmentosa. Moreover, we offer a translational strategy for accurate WGS-data prioritization, which is essential for the advancement of personalized medicine.

Expanding the phenotype of THRB: a range of macular dystrophies as the major clinical manifestations in patients with a dominant splicing variant

“…In fact, as the THRB gene was already included in the targeted diagnostic panel for RTHβ patients, we searched for the presence of likely pathogenic variants in all coding exons and its splice junctions of this gene in the unsolved IRD cohort. Prioritization of THRB variants was done using the prediction tools and optimized cutoffs previously described elsewhere (Gonzalez-Del Pozo et al, 2022).…”

“…The WGS data prioritization was conducted using our validated pipeline as previously described (Gonzalez-Del Pozo et al, 2022). Briefly, the WGS data of the four sequenced individuals were combined using VCF combine (Vcflib).…”

“…The application of two additional pathogenic ACMG/AMP rules, PP1 (cosegregation with the disease in multiple affected family Variant filtering scheme used for the prioritization of the NGS data. (A) The starting point was the combined vcf file from the four sequenced individuals of family A (FamA-II:1, II:2, III:1, and III:2) that was filtered using the settings and the cutoffs described elsewhere (Gonzalez-Del Pozo et al, 2022). Briefly, variants with a MAF>0.01 in gnomAD database were filtered out except those variants classified as pathogenic (P), likely pathogenic (LP), or as conflicting interpretations of pathogenicity with P or LP entries (CIP: P/LP) in ClinVar database.…”

“…However, the overlapping causative genes and phenotypes tremendously complicate the genetic analysis of IRDs, and a relatively large percentage of affected individuals (~40%) remain genetically unsolved after routine analyses (Martín-Sánchez et al, 2020). In this sense, the application of a global approach like whole genome sequencing (WGS) promises to increase the diagnostic yield of IRDs through the identification of challenging variants like structural variants, deep-intronic variants, or variants in novel causative genes (Gonzalez-Del Pozo et al, 2022).…”

A Comprehensive Evaluation of the Performance of Prediction Algorithms on Clinically Relevant Missense Variants

“…Over the last two decades, a multitude of computational tools have been developed to address this emerging issue. These prediction algorithms are widely used as evidence to prioritize and select novel variants of unknown significance for in vivo and in vitro functional assays [ 6 ]; they are integrated into NGS bioinformatics pipelines [ 7 ], and some of them, such as Polyphen2 [ 8 ], SIFT [ 9 ], and MutationTaster [ 10 ], are integrated into commercially available interfaces and are routinely used in clinical diagnostics [ 11 ].…”