Table 1.

Detailed breakdown of Illumina sequences generated from exon catches

Length and kind of Illumina sequencing reads	36-base GA-I end sequences	36-base GA-I shotgun sequences	76-base GA-II end sequences
Aggregate length of targeta	2.5 Mb	2.5 Mb	2.5 Mb
Aggregate length of baits	3.7 Mb	3.7 Mb	3.7 Mb
Total raw unfiltered sequence	152 Mb	219 Mbb	851 Mb
Raw sequence not aligned uniquely to genomec	67 Mb	116 Mb	358 Mb
Uniquely aligned human sequence	85 Mb	102 Mb	492 Mb
Uniquely aligned sequence on target	36 Mb	51 Mb	235 Mb
Uniquely aligned sequence near targetd	40 Mb	38 Mb	210 Mb
Uniquely aligned sequence on or near target	76 Mb	90 Mb	445 Mb
Fraction of uniquely aligned sequence on or near targete	89%	88%	90%
Fraction of raw bases uniquely aligned on or near targetf	50%	41%g	52%
Fraction of uniquely aligned bases on targeth	42%	50%	48%

^aProtein-coding exon sequence only

^bEach unit of concatenated catch contains 44–46 bases (~18%) of generic adapter sequence. Therefore, ~18% (39 Mb) of the 219 Mb is not of human origin.

^cAll raw sequence that fails to align uniquely to the human reference genome including low-quality sequence

^dOutside but within 500 bp of a target exon

^eUpper bound for estimating the specificity of hybrid selection

^fLower bound for estimating the specificity of hybrid selection

^gThe denominator (219 Mb) includes ~39 Mb of sequence from the generic adapters. Excluding these 39 Mb, the lower bound for the estimated specificity of this catch is 90/180 = 50%.

^hUpper bound for the overall specificity of targeted exon sequencing