Shed skin as a source of DNA for genotyping-by-sequencing (GBS) in reptiles

Abstract

Over a fifth of reptile species are classified as ‘Threatened’ and conservation efforts, especially those aimed at recovery of isolated or fragmented populations, will require genetic and genomic data and resources. Shed skins of snakes and other reptiles contain DNA; are a safe and ethical way of non-invasively sampling large numbers of individuals; and provide a simple mechanism by which to involve the public in scientific research. Here we test whether the DNA in dried shed skin is suitable for reduced representation sequencing approaches, specifically genotyping-by-sequencing (GBS). Shed skin-derived libraries resulted in fewer sequenced reads than those from snap-frozen muscle samples, and contained slightly fewer variants (70,685 SNPs versus 97,724), but this issue can easily be rectified with deeper sequencing of shed skin-derived libraries. Skin-derived libraries also have a very slight (but significantly different) profile of transitions and transversions, most likely as a result of DNA damage, but the impact of this is minimal given the large number of single nucleotide polymorphisms (SNPs) involved. SNP density tends to scale with chromosome length, and microchromosomes have a significantly higher SNP density than macrochromosomes, most likely because of their higher GC content. Overall, shed skin provides DNA of sufficient quality and quantity for the identification of large number of SNPs, but requires greater sequencing depth, and consideration of the GC richness of microchromosomes when selecting restriction enzymes.