Table 2.

Comparison of genotyping approaches

	Cost per samplea	Cost per marker data pointa	SNP discovery rate	Analysis complexity	Prior genomic knowledge	Preferred population type	Drawbacks	Applications
RADseq	Low	Moderate	Low to moderate	Moderate	No	All	Labour‐intensive library preparation; high read depth variation	De novo SNP discovery, genome improvement, genetic mapping
Elshire GBS	Low	Moderate	Low	Moderate	No	All	High levels of missing data	De novo SNP discovery in simple genomes, genome improvement, genetic mapping
ddRAD	Low	Moderate	Low to moderate	Moderate	No	All	Sensitive to allele dropout; high‐quality sample required	De novo SNP discovery, genome improvement, genetic mapping
Parental inference WGR	High	Low	High	High	No	Biparental cross	High cost; inference is error‐prone	De novo SNP discovery, high‐resolution mapping of (complex) plant genomes, genome improvement
SkimGBS	High	Low	High	High	Yes	Biparental cross	High cost; need for prior genomic information	SNP discovery and high‐resolution mapping of (complex) plant genomes, genome improvement
SNP array	Moderate	High	High	Low	Yes	All	Ascertainment bias; need for prior genomic information	SNP discovery and high‐resolution mapping, genetic mapping
Exome sequencing	Moderate	High	Low	Moderate	Yes	All	Need for prior genomic information	SNP discovery in complex genomes, genetic mapping
RNA‐seq	Moderate	High	Low	Moderate	No	All	Biases in transcript abundances	SNP discovery in complex genomes, genetic mapping, expression analysis

^aRelative costs shown may vary with factors including sample number and target coverage.