Extended Data Fig. 1: Overview of filtering strategy to remove potential bioinformatic artifacts of copy-number variation (CNV) in major venom gene regions.

a Schematic representation of the workflow used to filter departures from Hardy–Weinberg equilibrium (HWE) due to excess heterozygosity at SNP positions. Input variant calls were analysed using a one-tailed HWE test (adapted from Wigginton et al. 2005) and SNPs with significant HWE P-values after FDR correction were filtered. Individual CNVs were detected, scored, and regional variation in CNV presence, quantified as the proportion of individuals in a population with a detected CNV (% CNV), was measured across each major venom gene region. CNV calls were used to mask genotypes per individual prior to population genetic analysis. b Regional variation in CNV presence per population across the major venom gene regions. Top panels for each venom gene region show log2 read depths in sliding windows relative to the autosomal median depth for the CV1 reference population. Here values of zero indicate equal coverage to the autosomal median, values of −1 equal half coverage, and values of 1 equal twice the autosomal median coverage. Dark blue segments indicate the locations of venom genes in each region. Lower panels for each region show variation in the proportion of individuals with masked genotypes in a detected CNV (% CNV) per population. c Minor allele frequency spectra across the dataset for each gene in the major venom gene families after using the described HWE and CNV filtering strategy.