A whole exome variant filtering software for identification of disease causing variants

Abstract

Purpose: To develop and validate a software program that can efficiently analyze sequence changes identified by whole exome and/or whole genome sequencing to detect causal variants.

Methods: exomeSuite is a freely available multi-platform application we have designed to analyze exomes of both single familial cases and large cohorts consisting of multiple pedigrees or a group of unrelated individuals. The software identifies candidate genes that exhibit sequence variants consistent with the predicted pattern(s) of inheritance. The candidate gene list can be further filtered for any number and combination of user specified criteria which are limited only by the information provided in the input files (e.g. gene minor allele frequency, absence or presence in dbSNP, and 1000 genomes, etc). Furthermore, we have integrated into the software the ability for users to filter the data by the annotation of the missense mutation predicted by PolyPhen, and the EST profile in the NCBI UniGene Database. Exome variants of three pedigrees with retinal degeneration were analyzed using exomeSuite.

Results: exomeSuite software filters data for sequence variations (SNVs and indels) following either dominant, recessive homozygous, or compound heterozygous pattern of inheritance for the monogenic disease of interest. This software can also be used for homozygosity mapping and for the identification of variants associated with genetic traits in a population. In addition, the user may define a list of “genes of interest” to filter results; the software by default includes a list of RetNet genes. Additionally, exomeSuite can build databases of sequenced exomes and filter results based on allele frequency within the database. Results will vary between pedigrees and based on the number of exomes sequenced, the inheritance pattern filtered for, and on user defined filters. Typically we have filtered over 50,000 variant calls to fewer than 10, in some instances just one or two. Analysis of the three pedigrees used to validate exomeSuite led to the identification of causative mutations in the USH2A, RPE65 and RBP4 genes.

Conclusions: exomeSuite rapidly enables users with little to no computer programming and/or bioinformatics experience to filter large datasets for variants segregating with disease, allowing virtually anyone to capitalize on the ever increasing throughput and decreasing cost of next generation sequencing technologies.

Keywords: 539 genetics • 696 retinal degenerations: hereditary • 537 gene screening