"Some special term seems desirable to direct attention to variation within groups, and I propose the word cline, meaning a gradation in measurable characters...Prefixes can be used to denote clines of different types, for example, genocline (gradient in genes), geocline (geographical cline), chronocline (paleontological trend), etc."
Julian Huxley, Clines: an Auxiliary Taxonomic Principle, Nature, 1938.
Applying the method of Coop et al. (2010) [PubMed], the Di Rienzo lab conducted a genome-wide study to identify regions associated with variation in 20 environmental variables across 61 worldwide populations (52 HGDP + 4 HapMap Phase III + 5 in house genotyped on the Illumina 650Y platform; Hancock et al. 2010 [PubMed], 2011 [PubMed]). The complete results of this analysis can be accessed through this database portal.
The complete set of Bayes factors can be separately downloaded for the continuous (climate and geography) variables [climate.4.tar.gz (89 MB)] and categorical (ecoregion and subsistence) variables [ecoreg_subsist.tar.gz (90 MB)]. The population allele frequency data from which the Bayes factors were calculated is also available [freq_data.tar.gz (80 MB)].
Note: The transformed rank statistic has changed and is now calculated in the following manner: Each SNP's Bayes factor is calculated by averaging the results from 5 independent runs of Bayenv. The SNPs are binned as described previously (Hancock et al. 2011 [PubMed]). A standardized score is computed for each SNP using the mean and standard deviation of the Bayes factors in the SNP's bin. Finally, the SNP's empirical p-value is calculated based on the rank of its standardized score relative to the entire set of tested SNPs.
|Filter results based on variable of interest and empirical p-value threshold.|
|Maximum Transformed Rank||Environmental Variable|
|Filter results based on minimum level of LD with queried HapMap SNP.|
|Minimum r2||HapMap Population|