Table 1

Comparison of four high-throughput polymorphism detection approaches



RAD tagging

RAD sequencing


Marker type

SNPs and indels

Restriction cut site polymorphisms

Sequence data: SNPs next to restriction cut sites

Restriction cut site polymorphisms: distinguishes SNPs and indels

Number of loci surveyed


19,200 (elements on an enriched RAD-tag microarray designed from stickleback)

26 nucleotides at 41,622 RAD tags


Number of polymorphisms identified (informative marker rate)

3,806 (4% at a 5% false discovery rate cutoff)

1,990 (10% at a two-fold signal difference cutoff)

Approximately 13,000 (31%)

12,431 (24%)

False discovery rate

3% (117 out of 121 confirmed correct by sequencing)

9% (20 out of 22 confirmed correct by sequencing)

Not reported

<1% (113 out of 114 confirmed correct by sequencing)


Custom high-density oligonucleotide array (Affymetrix), 25 bp oligo

cDNA or genomic tiling array (in house synthesis)

Illumina sequencing

Custom high-density oligonucleotide array (Agilent), 50 bp oligo

Prior information required

EST, 454 or genome sequence

EST or RAD-tag library for array synthesis

EST or genome sequence to map short sequence reads

EST, 454 or genome sequence

Polymorphism identification

Hybridization signal difference among study individuals

Hybridization signal difference between two study individuals

Custom Perl scripts for sequence alignment

Genotype clusters across all study individuals

Individual genotype data





Organisms studied

Yeast, Arabidopsis, Anopheles, several seed plantsa

Drosophila, stickleback, zebrafish, Neurospora


Purple sea urchin

Numbers are from studies that describe each method: SFP [26]; RAD tagging [25]; RAD sequencing [50]. aSee Gupta et al. [23] for review of high-throughput applications in crop plants.

Pespeni et al. Genome Biology 2010 11:R44   doi:10.1186/gb-2010-11-4-r44

Open Data