Development of optimal wet-lab and bioinformatics protocols for implementation of RAD sequencing for NERC science

RAD ABSTRACT A common problem in ecological genetics is the development and deployment of markers in wild populations. This can be a time-consuming and expensive task for non-model organisms, and can be a serious block to achieving research goals. Recently, restriction-site associated DNA sequencing (RADSeq) has emerged as a technology with the potential to simultaneously discover, validate and score robustly a large number of markers (in the thousands) across any genome. Using ultra-high throughput sequencing technologies it is possible to screen hundreds of individuals per week. Two issues hinder wide take-up of this technology: the difficulties encountered in preparing samples for RADSeq, and the analysis of the tens to hundreds of millions of sequence data points generated. Here we propose to establish in the GenePool genomics facility (a collaborating centre in NERC's NBAF) the resources and know-how to deliver this game-changing technique to UK environmental and population genetics research. We will develop best-practice, streamlined and repeatable laboratory methods that will deliver -robust methods for generating RADSeq libraries from large sample sizes, -optimisation of the application of molecular indexing (to permit multiplexing), -proof of the use of multiple different restriction enzymes (sampling independent populations of sites) and -systems for robust paired-end sequencing (to scan longer regions per RAD site for SNPs). There are no validated software tools for analysis of RADSeq data, and our exploration of the small datasets we have developed in house suggest that patterns of error in sequences and differential representation of sites in datasets makes data processing non-trivial. We will build easy-to use pipelines for RADSeq data analysis, incorporating best-practice quality checking, error management and outputs ready for further analyses in third party software. These pipelines will be used to verify the mapping of dauer entry and other traits in the C. elegans model system, and to deliver genetic analysis of RAD sites in the other genomes. These tools and protocols will subsequently be offered in-house to NERC science, and also disseminated through training and publication. We will also make available the validated RADSeq adapter sets at cost to NERC science. We will use three test systems. The major testbed will be a set of recombinant inbred lines derived from, and newly constructed crosses between, wild strains of the nematode Caenorhabditis elegans, where we will investigate the use of RADSeq markers in fine mapping of traits in a fully-sequenced genome. We will also construct test libraries from two other organisms, the oak Quercus robur and the burying beetle Nicrophorus vespilloides, to examine RADSeq in larger genomes, and in organisms with no genome data existing.