Rare-variant and polygenic analyses of amyotrophic lateral sclerosis in the French-Canadian genome

Purpose: The genetic etiology of amyotrophic lateral sclerosis (ALS) includes few rare, largeeffect variants and potentially many common, small-effect variants per case. The genetic risk liability for ALS might require a threshold comprised of a certain amount of variants. Here, we tested the degree to which risk for ALS was affected by rare variants in ALS genes, polygenic risk score, or both. Methods: 335 ALS cases and 356 controls from Que ' bec, Canada were concurrently tested by microarray genotyping and targeted sequencing of ALS genes known at the time of study inception. ALS genome-wide association studies summary statistics were used to estimate an ALS polygenic risk score (PRS). Cases and controls were subdivided into rare-variant heterozygotes and non-heterozygotes. Results: Risk for ALS was significantly associated with PRS and rare variants independently in a logistic regression model. Although ALS PRS predicted a small amount of ALS risk overall, the effect was most pronounced between ALS cases and controls that were not heterozygous for a rare variant in the ALS genes surveyed. Conclusion: Both PRS and rare variants in ALS genes impact risk for ALS. PRS for ALS is most informative when rare variants are not observed in ALS genes.