Genome Coverage vs Sequencing Depth: How Sequencing Parameters Affect Detection of Population Structure

Next generation DNA sequencing of thousands of single nucleotide polymorphisms is rapidly replacing traditional genetic markers in the field of molecular ecology. The resulting tremendous increase in data has enhanced many aspects of molecular ecology, including analysis of population structure, local adaptation and phylogenomics. Ideally, researchers would sequence entire genomes, but the cost of sequencing most eukaryotes is prohibitive. Reduced representation sequencing approaches such as RADSeq, ddRADSeq and GBS allow researchers to sequence a small, specific subset of homologous regions of the genome across individuals, drastically reducing the cost of sequencing. However, most labs are still limited in the amount of sequencing output they are able to generate. Consequently, most molecular ecologists face a tradeoff between sequencing depth and genomic coverage, most of the time without a priori knowledge regarding which aspect of sequencing is more important. To address this knowledge gap, we used nextRAD to sequence a reduced representation library from n = 143 lake whitefish in Lake Huron that showed weak population structure using microsatellites. We sequenced the same individuals at both high and low sequence coverage: >15X and >3X, respectively. As a result of fixed output, individuals sequenced at low depth have higher genome coverage (>50,000 loci), compared to those sequenced at high depth (>20,000 loci). In this poster we compare analyses of population structure using these two sequencing datasets. Our results will help researchers decide how to handle the depth vs coverage tradeoff in future studies.