Abstracts

Rationale: Epilepsy susceptibility has a clear genetic component, but causative mutations have only been identified in a handful of genes and account for a very small number of cases. Consistent with a complex disease, epilepsy exhibits extreme locus heterogeneity, making more traditional approaches such as GWAS and linkage studies difficult and often unsuccessful. We hypothesize that rare, highly penetrant mutations are likely to be found in families containing multiple affected individuals, and hence whole-genome sequencing (WGS) in affected individuals from such families can be used for discovery of pathogenic variants. Methods: To test this hypothesis we performed WGS in two affected individuals from each of 9 families containing multiple individuals with non-acquired epilepsy (average 6.2 affected per family). To reduce the number of variants shared by chance, we selected the two most distantly related affected individuals from each family. The Illumina GAIIx was used to sequence each of the 18 genomes, which were sequenced to an average coverage of ~31x. The short sequence reads were aligned to the reference genome using BWA software. Two main types of variants, single nucleotide variants and small insertion-deletions, were identified using SAMTools software. We annotated and analyzed all identified variants using Sequence Variant Analyzer, an in-house software program. Within each family, the genomes of affected relative pairs were analyzed simultaneously to identify rare shared functional variants (i.e., nonsynonymous, protein truncating, or splice site-disrupting variants). Variants were considered to be rare if they were present at a frequency lower than the cumulative incidence of epilepsy in the general population (~3%) based on comparison to HapMap and a set of 80 whole-genome or whole-exome sequenced controls from our laboratory. Results: No rare functional variants were present in all 18 epilepsy genomes, and no shared functional variants were found in any of 12 genes already known to play a role in Mendelian forms of epilepsy. Analysis of sharing of rare functional variants in distantly-related affected individuals reduced the set of candidate variants to a manageable number for genotyping in larger cohorts. The number of rare, shared functional variants averaged 108 (min 49, max 201) per family. Very few shared variants were found in multiple families: 17 were found in 2 families, 2 in 3 families, and 1 in 4 families. This suggests that different genetic factors are responsible for epilepsy in most of the families. This work defines a set of 974 candidate variants that can be followed up by genotyping in larger cohorts and in other members of the same families. Conclusions: We demonstrate that WGS of distantly-related affected individuals in multiplex epilepsy families is a powerful strategy for identifying potentially pathogenic candidate variants. Analysis and additional validation, including cosegregation of candidate variants in the extended pedigrees, is on-going.