A general linear model-based approach for inferring selection to climate

BMC Genomic Data

Table 1 Count of hits for each climate variable

Table 1a. Count of hits (-log10P > =5) for each climate variable partitioned by whether significance testing was conducted by scrambling globally across all populations (CS = 0) or with scrambling restricted to within each continent (CS = 1). Table 1b. Dependence of number of hits on the method of randomization used in significance testing. TOTAL gives the number of genic windows that achieve the specified significance, determined using CS=0 (worldwide scrambling) or CS=1 (scrambling within continents). Numbers in the table indicate overlap between the methods with the expected overlap in brackets, calculated assuming complete independence. Thus, the bottom left cell indicates that 147 of 328 hits achieving top significance with CS=0 achieved at least the lowest level of significance with CS=1 when the expected number was 15. Of these 147, 44 achieved at least P<=0.0001, including 16 that achieved top significance.

ISSN: 2730-6844