Comparing fragment size with read-length and their effect on HLA genotyping by next-generation sequencing (NGS)

Abstract

Aim We have previously studied the effect of shorter DNA fragments (less than 600 bp) on HLA typing by NGS of HLA-A, B, C, DRB1 and DQB on the Illumina MiSeq and found that 100–300 bp fragments produce more ambiguous typing results. Here, we evaluated the same reagents but altered the conditions to produce longer DNA fragments in order to determine if the number of ambiguous typing results could be reduced. We have previously shown that Illumina 2 × 250 sequencing resulted in less ambiguity than 2 × 150 bp sequencing. Here, we asked if further increasing the sequencing read length (2 × 300 bp) leads to additional gains in accuracy. Methods Ten IHWG samples had long-range PCR performed for HLA-A, B, C, DRB1 and DQB1 using a combination of in-house and commercial primers. After PCR clean-up and quantification, each sample’s 5 loci were pooled. Using the NEBNext dsDNA Fragmentase kit and various incubation times, the long-range PCR products were digested to produce 600–1200 bp fragments size selected with magnetic beads (AMPure Beads, Beckman Coulter) or the Blue PippinPrep (Sage Science). The NEBNext Library Prep Kit was used to complete library preparation. Unique indices were ligated onto each sample. Each sample library was measured for size and quantified. Next, all sample libraries were equimolarly pooled and two Illumina Reagent kits were run, v2 (500 cycle) and v3 (600 cycle), each loaded at a 10 pM concentration. Fastq files were analyzed using Omixon’s Target v.1.8.1 software for HLA genotyping. Results Ninety-six percent of the IHWG samples, each with 5 loci and an average fragment size of 684 ± 73 bp, typed correctly without any ambiguity. Samples with an average fragment size of 973 ± 132 bp performed the same, 96%. There was no significant difference in results between the 2 × 250 and 2 × 300 read length kits using samples with shorter (684 bp) or longer (973 bp) fragment sizes. Conclusions We have now compared fragment sizes from 100–300 bp, to 900–1200 bp, and found that generally, fragment sizes 600 bp or greater produce the most correct and unambiguous HLA typing assignments. We have previously determined that 2 × 250 read length kit performs better than the 2 × 150 kit, and now we show that the 2 × 250 read length kit is preferred over the 2 × 300 bp kit due to its shorter run time (40 h vs. 60 h) with equal performance.