LAUSR

Aim This study examined the variation observed when typing exons 2 + 3 of HLA-DQB1 using reverse sequence specific oligonucleotides (rSSO) as compared to typing exon 2 alone in volunteers from an unrelated hematopoietic stem cell donor program. New alleles and those with incomplete sequences were characterized with next generation sequencing (NGS) to update the IPD-IMGT/HLA database. Methods Samples were typed for HLA-DQB1 using One Lambda’s SSO LabType with additional phasing probes for exon 2 ( ∼ 50,000 donors) or for exons 2 + 3 ( ∼ 15,000). Full-length HLA-DQB1 sequences (40) were obtained using an Illumina platform and in-house protocol. Results Overall, 38 incomplete DQB1 alleles were identified and sequenced with NGS, including 6 that could only be detected by rSSO with exon 3 probes. The majority of these alleles appear only 1–2 times in the sample set and were not listed as common or well-documented in Mack et al. 2013. The addition of exon 3 in rSSO typing complicated analysis since many alleles lack sequence data for exon 3, and the software assumes no probe binding if there is no known sequence. An example is DQB1*06:08 ( ∼ 10), which binds to one of the exon 3 probes and therefore is not assigned (i.e., false positive). Two alleles showed discrepant results between rSSO and NGS due to novel sequences. The full-length sequences obtained for all alleles sequenced demonstrate high conservation of the regions outside of exon 2. Conclusions There are advantages and disadvantages from screening exons 2 + 3 of HLA-DQB1. It elucidates greater variation not identified from screening exon 2 alone but has a more complicated analysis due to incomplete DNA sequences. Updating the IPD-IMGT/HLA database is critical for improved accuracy. C. Masaberg: 7. Other (Identify); Company/Organization; Intellectual property typing workflow, One Lambda licensing. C.K. Hurley: 7. Other (Identify); Company/Organization; Intellectual property typing workflow, One Lambda licensing.