Table 2. Overview of studies evaluating exome enrichment platforms as well as summary of which of the platforms performed best for the assessed aspects.
This study | Clark et al. 2011 (3) | Asan et al. 2011 (2) | Parla et al. 2011 (4) | Sulonen et al. 2011 (5) | Chilamakuri et al. 2014 (6) | |
---|---|---|---|---|---|---|
Enrichment platforms | Agilent v5+UTR, NimbleGen v3+UTR and Illumina Nextera Expanded Exome | Agilent v3, NimbleGen v2 and Illumina TruSeq Exome | Agilent v1, NimbleGen v1 (in-solution), 2.1M array | Agilent v1 and NimbleGen v1 | Agilent v1, v3 and NimbleGen v1, v2 | Agilent v4, NimbleGen v3, Illumina TruSeq Exome and Illumina Nextera Expanded Exome |
Sequencing platform | Illumina HiSeq 2000/2500 paired-end 100-bp reads | Illumina HiSeq 2000 paired-end 100-bp reads | Illumina HiSeq 2000 paired-end 90-bp reads | Illumina GAIIx, paired-end 76-bp reads | Illumina GAIIx, paired-end 82-bp reads | Illumina HiSeq 2000 paired-end 100-bp reads |
DNA samples | Six samples performed by different vendors, 24 samples performed by one vendor using Agilent | One sample | One sample | Six HapMap samples (two for both platforms and four only for NimbleGen) | One sample for all platforms, 25 samples for one platform | One sample with two technical replicates per platform |
Region for sequence variant calling | Common designed target region in RefSeq coding exons 100% covered at 20× by all platform-vendor combinations | Genome-wide | Designed target region with 200-bp flanking sequences | CCDS | Genome-wide, designed target region (individual and common), and CCDS | Designed target region (individual and common), CCDS, RefSeq (coding and UTR) and Ensembl |
Largest designed target region | NimbleGen | Illumina | Agilent | Agilent | Agilent v2 | NimbleGen |
Largest coding region (reference) | NimbleGen (RefSeq) | Agilent (RefSeq, Ensembl CDS) | Agilent (CCDS) | Agilent (CCDS) | Agilent v2 (CCDS) | Illumina (CCDS, RefSeq, Ensembl) |
Best designed target enrichment efficiency | Agilent | NimbleGen | NimbleGen (array and in-solution) | NimbleGen | NimbleGen v2 | Agilent |
Lowest off-target enrichment | Agilent and NimbleGen | NimbleGen | NimbleGen (array and in-solution) | NimbleGen | NimbleGen v1 | Agilent and NimbleGena |
Best GC-rich region enrichment | Agilent | Agilent | NimbleGen array | No data | NimbleGen v2 | Illumina Nextera |
Highest accuracy of SNV detection (benchmark) | Agilent (Sanger sequencing, MLPA and SNP array) | Agilent (SNP array) | No clear difference among platforms (SNP array and WGS) | Agilent (HapMap and 1000 Genome Project data) | NimbleGen v2 (SNP array) | No determination of accuracy by comparison to a benchmark (only calling of SNVs) |
aEstimated from provided figures, as off-target reads were reported as relative proportion of filtered reads rather than total mapped reads; CCDS, Consensus Coding Sequences.