Evidence of Polygenic Adaptation in Response to Anthropogenic Water Contamination in North Atlantic Eels

According to the quantitative genetic theory, a rapid adaptation of complex traits occurred through subtle variation in allelic frequencies on several covarying loci. In genomics studies, gene-scan is the common approach to detect loci under selection, however this method does not search for correlation among loci. The random forest algorithm (a tree-based ensemble machine learning tool) appears to be more suited to detect evidence of polygenic adaptation since takes into account the correlation and interactions among loci. In this study, single-nucleotide polymorphism obtained via genotype-by-sequencing were used to assess the presence of a polygenic adaptation in both endangered North Atlantic eels (Anguilla anguilla and A. rostrata) in response to anthropogenic water contamination, using the random forest algorithm. Four populations (2 in contaminated and 2 in clean sites) for each species were sampled for 25 individuals each. Ultimately, this research will provide a better understanding of the genetic basis behind the complex adaptation in contaminated water environments and will improve eels and other fish conservation program.