P097 An HLA typing assay using the illumina NGS system allows for the detection of poorly amplified HLA alleles, virtually eliminating allele dropout

Abstract

Aim DNA sequencing has long been the gold standard for HLA typing, offering the ability to interrogate each base pair of a targeted region. Sanger sequencing-based typing assays are known to suffer from allele dropout, whereby one of two encoded alleles is less represented in the results, due to poor amplification. Next-generation sequencing (NGS) includes several distinct technologies, defined by their ability to sequence clonal DNA fragments in a high-throughput manner. One such technology, provided by Illumina, sequences thousands of short DNA fragments, with read depth orders of magnitude higher than Sanger sequencing, which may allow for the identification of poorly-amplified alleles. Here we sought to determine the rate of allele dropout in an Illumina-based HLA typing assay. Methods NGSgo (GenDx) was used to produce libraries for the HLA-A, B, C, DRB1, DRB3/4/5, DQA1, DQB1, DPA1, and DPB1 loci for sequencing on a MiSeq (Illumina). NGSengine (GenDx) was used to analyze the data and type the loci. 85 unique samples were run in parallel by a combination of previously-validated assays (SBT, SSP, rSSOP) and by NGS. Results The NGS assay resulted in very high read depths (>100 for our clinical results) and a very low error rate. Use of NGSengine software allowed detection of both alleles at heterozygous loci when the alleles were unevenly represented (imbalanced). In cases of extreme allele imbalance, data visualization by NGSengine allowed detection of the minor allele during manual review. Of the 811 patient loci tested by the previously-validated assays, 584 were heterozygous, 108 were homozygous and 119 were absent; the NGS results were 100% concordant. Conclusions In our experience with a clinical NGS typing assay, we found the Illumina sequencing technology was able to successfully identify all alleles present in tested samples, including poorly-amplified alleles. While the possibility of rare allele dropout still exists with this technology, due to alleles with novel polymorphisms at a primer-binding site, we found that with an appropriate combination of NGS library construction, sequencing technology, and analysis software, allele dropout is a low frequency, to non-existent, event. This reduces the need for extra testing to confirm homozygous loci during routine clinical HLA typing.