Large, three-generation human families reveal post-zygotic mosaicism and variability in germline mutation accumulation.

t-PGCS DNMs and likely post-zygotic DNMs) we first fit a simple Poisson regression model that calculated the effect of paternal age on total autosomal DNM counts in the R statistical language (v3.5.1) as follows: glm(autosomal_dnms ~ dad_age, family = poisson(link='identity’)) This model returned a highly significant effect of paternal age on total DNM counts (1.72 DNMs per year of paternal age, p<2e-16), but was agnostic to the family from which each third-generation individual was ‘sampled.’ Importantly, a number of third-generation individuals in the CEPH/Utah cohort share grandparents, and may therefore be considered members of the same family, despite having unique second-generation parents ( Figure 3—figure supplement 1 ).

also in 132,142 other papers