It’s been a few years since the Hoekstra lab at Harvard first published its double-digest RAD-seq protocol. Since then, the approach has been rapidly adopted by the community for massively parallel genotyping, particularly of non-model organisms, and has been the foundation for lots of new protocol and tool development.
ddRAD-seq was itself an iteration on the original RAD-seq (restriction-site associated DNA sequencing) method from the Cresko lab at the University of Oregon. It introduced an important tool for people interested in using the second digest step, which relied on automated size selection with the Pippin Prep to generate useful results.
In addition to the widespread use of ddRAD-seq for organisms ranging from fish to mosquitos, the community has continued to develop and expand RAD-based protocols. The recently published hyRAD protocol adds a hybridization capture step to make the approach useful with degraded DNA samples, such as those found in museum collections. EpiRADseq swaps in a methylation-sensitive restriction enzyme, enabling scalable and cost-effective quantification of methylation across whole genomes. And SimRAD is a novel software tool designed to improve ddRAD-seq results by accurately estimating the number of loci generated. Meanwhile, scientists have optimized ddRAD-seq methods for Ion Torrent sequencing and reviewed best practices for RAD-based approaches in general. One team compared ddRAD-seq to sequence capture and found that the RAD-based method generated more data for less money, noting that it would be particularly valuable for organisms without existing genome resources. This paper compared sequence data to SNP data from ddRAD-seq projects for phylogenetic inference, yielding advice for making phylogenetic trees from SNP data more accurate.
For a look at some great findings from recent ddRAD-seq studies, check out these papers: