Effect of diversity and missing data on genetic assignment with RAD-Seq markers

Reduced representation libraries are being used as a preferred source of markers to address population genetic questions. However, libraries of RAD-Seq variants often suffer from significant percentage of missing data. In addition, algorithms used to mine SNPs from the raw data may also underscore biological variation. We investigate the effect of biological diversity in mining SNPs from the program STACKS and the effect of missing data on individual assignment implemented in STRUCTURE. We observed that changing diversity parameters in STACKS significantly alters the number of SNPs discovered and allowing for higher percentage of missing data retrieves more loci and possibly more power for individual assignment.