Paris (front and centre) leading a workshop on how to analyse RAD-seq data
New RAD-seq Revolution: New Method Optimises Analysis
An international partnership between the University of Exeter and University of Illinois at Urbana–Champaign have recently published the first method of its kind for standardising analysis of Restriction site-Associated DNA sequencing (RAD-seq) data.
RAD-seq is a relatively old sequencing technology (first originating in 2008); whereby only a small subset of a genome is sequenced in order to give a representative sample of the whole genome. Since only a small fraction of the whole genome is sequenced RAD-seq is vastly cheaper and faster than traditional whole-genome sequencing. Because of this, the technique has grown to prominence within the field of population genetics as it allows many individuals of a species to be sequenced and compared to indicate genetic differences that may occur within and between populations.
Although RAD-seq has been around for a while, thus far there has been very little standardisation when it comes to analysing the data with many papers having been published presenting sub-optimal assemblages.
Josephine Paris, the lead author on the study, said “This research arose from several years of interacting with other researchers using RAD-seq. It became apparent that a straightforward parameter optimisation strategy was lacking and researchers were concerned about confidently building these large RAD-seq datasets.”
Using RAD-seq data from populations of brown trout, king penguins and red earthworms, Paris presents figures showing that each dataset required different parameters in order to generate the optimum assemblies from the raw data.
Subsequently the team has outlined a new method which can be used to generate optimal parameters for the dataset from scratch regardless of the origin of the dataset.
Ms Paris hopes “[the new method] will provide a vital ‘road map’ for researchers embarking on RAD-seq analyses for population genomics and phylogenomics”
The paper also outlines the issue with the current reliance on reference genomes when generatinggene assemblages by mapping raw reads directly onto a reference genome. The team provide evidence to suggest that it may actually be more accurate to generate loci de novo prior to integrating with the reference genome using the method they have developed.
The paper has been published in the journal Methods in Ecology and Evolution; Lost in parameter space: a road map for STACKS
Article written by Will Davison, Biosciences PressGang
Date: 24 February 2017