To compare the observed distribution of de novo coding variants per individual to the corresponding expected Poisson distribution we determined the frequency of the counts per individual for each cohort (TIC Genetics, TSAICG, and SSC Siblings) and then plotted this as a histogram. We next plotted a Poisson distribution using the dpois R function with lambda (λ) equal to the mean of the counts per individual. l was determined per cohort (see #2 below). All three cohorts appear to follow the expected Poisson distribution. However, to confirm this, we conducted a Chi Square goodness-of-fit test between the observed and expected distributions with the following steps (example R code in italics): Determine the number of individuals with 0, 1, 2, 3, 4, 5, or more de novo coding variants. Note that because during quality control we trimmed individuals with > 5 de novo variants the number of individuals with > 5 de novo coding variants is 0 for each cohort.counts = count(< df with number of de novos per individual >, numPassingCoding)$nCreate a vector of the number of de novo coding
