Performance comparison of whole-genome sequencing platforms

Abstract

Whole-genome sequencing is becoming commonplace, but the accuracy and completeness of variant calling by the most widely used platforms from Illumina and Complete Genomics have not been reported. Here we sequenced the genome of an individual with both technologies to a high average coverage of ∼76×, and compared their performance with respect to sequence coverage and calling of single-nucleotide variants (SNVs), insertions and deletions (indels). Although 88.1% of the ∼3.7 million unique SNVs were concordant between platforms, there were tens of thousands of platform-specific calls located in genes and other genomic regions. In contrast, 26.5% of indels were concordant between platforms. Target enrichment validated 92.7% of the concordant SNVs, whereas validation by genotyping array revealed a sensitivity of 99.3%. The validation experiments also suggested that >60% of the platform-specific variants were indeed present in the genome. Our results have important implications for understanding the accuracy and completeness of the genome sequencing platforms.

Figure 1. Genome coverage at different read depths

(a) Percentage of genome covered by different read depths in different platforms. (b) Histogram of genome coverage at different read depths.

Figure 2. SNV detection and intersection

(a) SNVs detected from the PBMC and saliva samples in each platform were combined. The unions of SNVs in each platform were then intersected. Sensitivity was measured against the Illumina Omni array. Ti/Tv is the transition-to-transversion ratio. The known and novel counts were based on dbSNP. ‘Sanger’ and ‘validated’ represent validation by Sanger sequencing and Illumina sequencing (with Agilent target enrichment capture), respectively. (b) Comparing platform-specific SNVs to non-SNV calls in another platform. IL, Illumina; CG, Complete Genomics.

Figure 3. SNV association with different genomic elements

(a) Gene elements: UTR, exonic, intronic and intergenic regions. Inset: number of SNVs associated with UTR5, UTR3 and exonic regions. (b) Gene elements: splicing sites, noncoding RNA and upstream/downstream (<1 kb) regions of genes. (c) Repetitive elements: centromere, telomere, tRNA and rRNA. (d) Repetitive elements: L1, Alu, simple repeat and low-complexity repeat. (e) SNV frequency at different chromosomal locations. Tracks from outer to inner: SNV frequency for Illumina (IL), Complete Genomics (CG), concordant, IL-specific and CG-specific calls. Outermost: chromosome ideogram.

Figure 4. Indel detection and intersection

(a) Indels detected from the PBMC and saliva samples in each platform were combined. The unions of indels in each platform were then intersected. Note: 5,668 IL and 8,415 CG indels were removed after 5b-window merging. (b) Indel size distribution. Negative size represents deletion and positive size represents insertion.