We hypothesized that batch effects might confound the combined analyses due to the use of three different exome capture arrays and sequencing at different centers (Table 1). Indeed, the three cohorts have different coverage distributions (Figure S3) and cluster separately in principal-component analysis (PCA) based on sequencing quality metrics (Figure S4). Likewise, we observed that “naïve” estimates of de novo variant rates were highly divergent across cohorts (Figure S5). However, we did not observe a significant difference in the “normalized” de novo variant rates between TIC Genetics, TSAICG, and the SSC control trios, suggesting that we adequately controlled for these confounds in our analyses.
