DArTseq generates two types of data:
- scores for presence/absence (dominant) markers, called SilicoDArTs as they are analogous to microarray DArTs, but extracted in silico from sequences obtained from genomic representations
- SNPs in fragments present in the representation.
It is also possible to extract Copy Number Variation (CNV) polymorphism information from some DArTseq representations.
The 0/1 scores are based on a range of DNA variation types: SNPs and small indels in restriction enzyme recognition sites, larger insertions/deletions in restriction fragments and at lower frequency, methylation variation at restriction sites when methylation sensitive enzymes are used in complexity reduction methods.
As most of DArTseq methods are using methylation sensitive RE the polymorphism patterns produced by DArTseq include a component of methylation profiling and therefore are capable of detecting epigenetic variation (similarly to most of microarray DArT methods). Compared to genome profiling using SNP assays only, the DArTseq method is much more comprehensive in terms of molecular variation underlying the polymorphisms and provides data at a more affordable price.
How DArTseq compares to other genotyping approaches using sequencing platforms?
The main difference between DArTseq and other genotyping by sequencing methods is how complexity reduction is achieved.
DArTseq complexity reduction methods are not only very simple (therefore high throughput and inexpensive), but also efficiently target low copy sequences. Often as many as 90% of DArTseq markers are aligning uniquely to a reference genome (for species with such resource available).
Deep sequencing of DArTseq representations ensures that the quality of extracted markers (both call rates and reproducibility) is very high. Finally, this deep sequencing enables reliable calling of heterozygotes in large proportion of SNP markers even in polyploidy species like wheat.