Table 2.

The number of SNPs (single nucleotide polymorphisms) that pass multiple test correction in the re‐analysed datasets. For Bonferroni correction the α threshhold 0.05 was divided by the total number of tests (SNPs tested) to get the genome‐wide multiple test correction threshold. For Q‐values and Benjamini–Hochberg (B–H) correction, a False Discovery Rate (FDR) threshhold of 0.05 was used. Bonferroni corrections carried out manually, B–H corrections followed the procedure in Benjamini and Hochberg (1995), and q‐values were calculated using the “qvalues” package in r (Storey et al., 2015)

Re‐analysis	Bonferroni	Q‐values	B–H
Raw counts	2	67,702	3,961
Counts scaled to n eff	3	67,505	4,571
Counts scaled to 100	456	61,022	15,053
Counts scaled to 1,000	33	13,013	2,532