Global inference of disease-causing single nucleotide variants from exome sequencing data

Whole-exome sequencing identifies novel pathogenic variants across the ATP7B gene and some modifiers of Wilson's disease phenotype

BACKGROUND & AIMS

Wilson's disease (WD) is an autosomal recessive disorder associated with disease-causing alterations across the ATP7B gene, with highly variable symptoms and age of onset. We aimed to assess whether the clinical variability of WD relates to modifier genes.

METHODS

A total of 248 WD patients were included, of whom 148 were diagnosed after age of 17. Human exome libraries were constructed using AmpliSeq technology and sequenced using the IonProton platform.

RESULTS

ATP7B p.His1069Gln mutation was present in 215 patients, with 112 homozygotes and 103 heterozygotes. Three other mutations: p.Gln1351Ter, p.Trp779Ter and c.3402delC were identified in >10 patients. Among patients, 117 had a homozygous mutation, 101 were compound heterozygotes, 27 had one heterozygous mutation, and 3 other patients had no identifiable pathogenic variant of ATP7B. Sixteen mutations were novel, found as part of a compound mutation or as a sole, homozygous mutation. For disease phenotype prediction, age at diagnosis was a deciding factor, while frameshift allelic variants of ATP7B and being male increased the odds of developing a neurological phenotype. Rare allelic variants in ESD and INO80 increased and decreased chances for the neurological phenotype, respectively, while rare variants in APOE and MBD6 decreased the chances of WD early manifestation. Compound mutations contributed to earlier age of onset.

CONCLUSIONS

In a Polish population, genetic screening for WD may help genotype for four variants (p.His1069Gln, p.Gln1351Ter, p.Trp779Ter and c.3402delC), with direct sequencing of all ATP7B amplicons as a second diagnostic step. We also identified some allelic variants that may modify a WD phenotype.

Leveraging multiple genomic data to prioritize disease-causing indels from exome sequencing data

The emergence of exome sequencing in recent years has enabled rapid and cost-effective detection of genetic variants in coding regions and offers a great opportunity to combine sequencing experiments with subsequent computational analysis for dissecting genetic basis of human inherited diseases. However, this strategy, though successful in practice, still faces such challenges as limited sample size and substantial number or diversity of candidate variants. To overcome these obstacles, researchers have been concentrated in the development of advanced computational methods and have recently achieved great progress for analysing single nucleotide variant. Nevertheless, it still remains unclear on how to analyse indels, another type of genetic variant that accounts for substantial proportion of known disease-causing variants. In this paper, we proposed an integrative method to effectively identify disease-causing indels from exome sequencing data. Specifically, we put forward a statistical method to combine five functional prediction scores, four genic association scores and a genic intolerance score to produce an integrated p-value, which could then be used for prioritizing candidate indels. We performed extensive simulation studies and demonstrated that our method achieved high accuracy in uncovering disease-causing indels. Our software is available at http://bioinfo.au.tsinghua.edu.cn/jianglab/IndelPrioritizer/.