A cost-and-time effective procedure to develop SNP markers for multiple species: A support for community genetics

Multispecies population genetics is an emerging field that provides insight relevant to conservation biology and community ecology. However, to date, this approach is limited to species with available genetic resources. The use of thousands of single-nucleotide polymorphism (SNP) markers developed from recent genotyping-by-sequencing (GBS) technologies is a roadmap for the study of nonmodel species, but remains cost prohibitive when several, distantly related species are involved. We aimed to overcome this issue using a single HiSeq3000 run of restriction-site associated DNA sequencing (RAD-Seq) to retrieve SNP markers for 40 diverse species including plants, invertebrates, fish, and mammals. We developed a Python-based pipeline to isolate c. 100–500 high-quality SNP markers for each species that could be genotyped through classical PCR amplification methods. To assess the quality of these markers, we validated our approach on c. 160 of the characterized SNPs for each of 18 Neotropical fish species from the river Maroni (French Guiana, South America), using the MassARRAY iPLEX platform from Agena Bioscience (San Diego, CA, USA). A run of the pipeline applying stringent filtering parameters enabled the successful design of between 130 and 3,492 SNP markers for 30 of the 40 study species. Relaxing pipeline parameters allows for an increase in the number of detected SNPs. Across the 18 species from French Guiana, an average of 85% of markers were successfully amplified, polymorphic, and scored in ≥90% of individuals (c. 200 individuals per species). The great majority (>98%) of these markers were at Hardy–Weinberg equilibrium in each sampling site from the river Maroni. This SNP discovery was performed at the cost of c. $US110 for each of the 40 species. Genotyping was performed at the cost of c. $US6000 for each of the 18 fish species with an average of 200 individuals per species. This strategy was found cost-and-time efficient to develop hundreds of SNP markers for a large range of nonmodel species, which can be used to investigate ecological and evolutionary questions that do not require whole-genome coverage