Genomic Datasets & Bioinformatics
The quality of the demultiplexed raw data for each sample was checked
using FastQC (v0.11.8; Andrews 2010) and summarized with MultiQC (v1.8;
Ewels et al. 2016; Table S1). Low quality sequences were discarded, and
the remainder were trimmed to uniform length (140bp) using theprocess_radtags program in STACKS (v2.5; Catchen et al. 2013).
The quality of the filtered datasets was then re-assessed using FastQC
and MultiQC (Table S2). Identification and genotyping of polymorphic
SNPs was conducted using the STACKS pipeline. We optimized the key
parameters M, n and m, following the protocol of Paris et al. (2017;
Appendix B). These parameters resulted in final datasets containing
2,494,940 and 2,325,766 polymorphic SNPs in S. umbilicalis andN. lapillus , respectively. The resulting datasets were filtered
to remove SNPs and samples of low quality and/or with high proportions
of missing data (Table 1; Appendix C), and file formats were converted
using PGDSpider (v2.1.1.5; Lischer & Excoffier 2012) for all downstream
analyses.