On the stability of the Bayenv method in assessing human SNP-environment associations

Table 1 Pairs of Bayenv runs compared

Letters indicate the parts of Tables 2 and 3 in which results for the indicated comparisons are averaged and presented. In Table 2, letters a, b, c, and d also correspond to parts e, f, g, and h, respectively. Hancock1 refers to the original Climate file posted on the dbCline website and published in Hancock et al. [40]; Hancock2 refers to file Climate.2 posted on dbCline website in autumn 2012 that uses the appropriate MCMC burn-in (see ‘Methods’ section); Blair1-5 refer to our five separate runs using 100,000 MCMC iterations; LongBlair1-5 refer to our five separate runs using 500,000 MCMC iterations; W/O_Sib1-5 refer to our five runs using a modified data set excluding the Siberian populations.

ISSN: 1479-7364