Genotyping and Quality Control of NAG-FIN Sample

Genotyping for the NAG-FIN sample was performed with the Human670-QuadCustom Illumina BeadChip (Illumina, San Diego, CA, USA) (N=1097) at the Wellcome Trust Sanger Institute, UK, and with the Illumina Human Core Exome BeadChip (N=901) at the Wellcome Trust Sanger Institute, and at the Broad Institute of MIT and Harvard, USA. Quality controls (QC) for the genotype data have been previously described^{21 (link)} and are also presented in Supplementary Table 1. Pre-phasing of the data was done with SHAPEIT2^{22 (link)} and imputation with IMPUTE2^{23 (link)} using the 1000 Genomes Phase I integrated haplotypes reference panel.^{24 (link)} For analyses of the 10 NSP genes (NRG1, NRG3, ERBB4, BACE1, PSEN1, PSEN2, APH1A, APH1B, PSENEN and NCSTN), we extracted SNPs within the gene regions (according to the longest isoform reported at the UCSC Genome browser) with 50 kb flanking regions. Gene boundaries (according to GRCh37/hg19) are listed in Supplementary Table 1, along with the number of SNPs included for each gene. Only variants with minor allele frequency (MAF) <0.01 located in coding regions, splice sites, promoters or untranslated regions (UTRs) were included in the rare variant analysis. Altogether, 15 036 SNPs were analyzed in our discovery phase.