Adapterama III: Quadruple-indexed, triple-enzyme RADseq libraries for about $1USD per Sample (3RAD)

Molecular ecologists frequently use genome reduction strategies that rely upon restriction enzyme digestion of genomic DNA to sample consistent portions of the genome from many individuals (e.g., RADseq, GBS). However, researchers often find the existing methods expensive to initiate and/or difficult to implement consistently, especially due to the inability to highly-multiplex samples to fill entire sequencing lanes. Here, we introduce a low-cost and highly robust approach for the construction of dual-digest RADseq libraries that relies on adapters and primers designed in Adapterama I. Major features of our method include: 1) minimizing the number of processing steps; 2) focusing on a single strand of sample DNA for library construction, allowing the use of a non-phosphorylated adapter on one end; 3) ligating adapters in the presence of active restriction enzymes, thereby reducing chimeras; 4) including an optional third restriction enzyme to cut apart adapter-dimers formed by the phosphorylated adapter, thus increasing the efficiency of adapter ligation to sample DNA, which is particularly effective when only low quantity/quality DNA samples are available; 5) interchangeable adapter designs; 6) incorporating variable-length internal indexes within the adapters to increase the scope of sample indexing, facilitate pooling, and increase sequence diversity; 7) maintaining compatibility with universal dual-indexed primers and thus, Illumina sequencing reagents and libraries; and, 8) easy modification for the identification of PCR duplicates. We present eight adapter designs that work with 72 restriction enzyme combinations. We demonstrate the efficiency of our approach by comparing it with existing methods, and we validate its utility through the discovery of many variable loci in a variety of non-model organisms. Our 2RAD/3RAD method is easy to perform, has low startup costs, has increased utility with low-concentration input DNA, and produces libraries that can be highly-multiplexed and pooled with other Illumina libraries.