Code and resulting candidate gene datasets from Anopheles genome environment association testing

The concept of a fundamental ecological niche is central to questions of
geographic distribution, population demography, species conservation, and
evolutionary potential. But robust inference of genomic regions associated
with evolutionary adaptation to particular environmental conditions
remains difficult due to the myriad of potential confounding processes
that can generate heterogeneous patterns of variation across the
genome.  Here, we interrogate the potential role of genome
environment association (GEA) testing as an initial step in building an
understanding of the genetic basis of ecological niche. We leverage
publicly available genomic data from the Anopheles gambiae 1000 Genomes
(Ag1000g) Consortium to test the ability of multiple, unique analytical
GEA methods to handle confounding genetic variation, control false
positive rates, and discern associations with broadly relevant climate
variables from randomly correlated allele frequency patterns throughout
the genome. We find evidence supporting the ability of commonly
implemented GEA methods to account for confounding patterns of spatial and
genetic variation, and control false positive rates. But we subsequently
fail to find evidence supporting the ability of GEA tests to reject
signals of adaptation to randomly simulated environmental variables,
indicating that discerning between true signals of genome environment
adaptation and genome environment correlations resulting from alternative
evolutionary processes remains challenging. Because signals of
environmental adaptation are so diffuse and confounded throughout the
genome, we argue that genomic adaptation to ecological niche is likely
best understood under an omnigenic model wherein highly interconnected,
genome-wide gene regulatory networks shape genomic adaptation to key
environmental conditions.