We performed simulations under 4 different scenarios: (i) using all chromosomes, (ii) using chromosomes 1-4, (iii) using chromosomes 1-2, and (iv) using chromosome 1 only. The motivation for performing simulations with a subset of chromosomes was to increase N/M, extrapolating to performance at larger sample sizes, as in previous work (Vilhjálmsson et al., 2015).
