P022NGS homopolymer detection error resulted A*03:01 mistyped as A*03:21n

Abstract

Aim Next Generation Sequencing (NGS) has enabled the characterization of complete HLA gene sequences and has become widely recognized as the preferred method for HLA typing. The importance of accurately adapting NGS into routine clinical testing has emerged. Homopolymer (HP) is a sequence of consecutive identical bases. Approximately 1.43 million HPs (also known as mononucleotide microsatellites) exist in the human exome, with the size of 4-mer and up. NGS have a relatively high error rate in determining HP due to the principles used for detection. Here we report an HLA sequencing error resulted from NGS HP detection limitation and hope to raise awareness and develop preventative steps. Methods: Routine clinical HLA allelic typing was performed using Immucor Mia Fora Flex kit. Stringent QC pass/fail criteria based on vendor’s recommendation were followed. Mia Fora Smart Flagging System Genotype status was incorporated into routine analysis. Two complementary bioinformatics, mapping and phasing, were used for analysis. Heterozygous calls were made if two nucleotides were present at a position with coverage ratio >20%. Results Two samples were typed as A*03:21 N during two NGS runs, A*03:21 N, A*33:01 and A*03:21 N, A*68:02 respectively. A*03:21 N is a well-documented allele, both assignments were acceptable by lab quality criteria. The frequency of A*03:21 N assignment caused concern and both samples were reflexed to Sanger SBT and confirmed as A*03:01. The difference is in exon 4, position 628, codon 186 where A*03:01 has 7-mer C vs. 8-mer C in A*03:21 N. A*03:21 N was defined as 627-628insC caused frameshift and premature stop at codon 196. Detailed NGS analyses revealed the number of ‘C’ call at position 628 were >20%, resulted A/C heterozygous calls and A*03:21 N assignment. Calling the common A*03:01 as non-expressed A*03:21 N has significant clinical impact. Even more problematic is that the assignments lacked warning sign. Conclusions NGS has revolutionized the field of genomic sequencing and rapidly adapted by HLA lab for clinical typing. However, NGS limitations, such as GC bias, preferential amplification and HP, are less known. Better understanding of these hidden risks and vendor software improvement can provide valuable benefit for labs utilizing NGS as their primary typing method.