Many human immunoglobulin heavy‐chain IGHV gene polymorphisms have been reported in error

Abstract

The identification of the genes that make up rearranged immunoglobulin genes is critical to many studies. For example, the enumeration of mutations in immunoglobulin genes is important for the prognosis of chronic lymphocytic leukemia, and this requires the accurate identification of the germline genes from which a particular sequence is derived. The immunoglobulin heavy-chain variable (IGHV) gene repertoire is generally considered to be highly polymorphic. In this report, we describe a bioinformatic analysis of germline and rearranged immunoglobulin gene sequences which casts doubt on the existence of a substantial proportion of reported germline polymorphisms. We report a five-level classification system for IGHV genes, which indicates the likelihood that the genes have been reported accurately. The classification scheme also reflects the likelihood that germline genes could be incorrectly identified in mutated VDJ rearrangements, because of similarities to other alleles. Of the 226 IGHV alleles that have previously been reported, our analysis suggests that 104 of these alleles almost certainly include sequence errors, and should be removed from the available repertoire. The analysis also highlights the presence of common mismatches, with respect to the germline, in many rearranged heavy-chain sequences, suggesting the existence of twelve previously unreported alleles. Sequencing of IGHV genes from six individuals in this study confirmed the existence of three of these alleles, which we designate IGHV3-49*04, IGHV3-49*05 and IGHV4-39*07. We therefore present a revised repertoire of expressed IGHV genes, which should substantially improve the accuracy of immunoglobulin gene analysis.

Automated analysis of high-throughput B-cell sequencing data reveals a high frequency of novel immunoglobulin V gene segment alleles

“…Given the large number of Rep-Seq studies being carried out on mRNA, these strict requirements mean that many novel alleles will not be represented in IMGT. This represents a significant challenge for Rep-Seq analysis with clinical implications (24), which could be avoided by adopting a classification system to indicate alleles with varying levels of underlying evidence, similar to what has been proposed (23). We are currently working to submit the novel alleles predicted by TIgGER to the UNSWIg human heavy chain repertoire (23), and they will also be hosted on our website (clip.med.yale.edu/tigger).…”

“…Eleven putative novel alleles (Table 2) were identified by applying the regression-based approach and filtering (described above) to the Ig sequencing data from seven subjects listed in Table 1. These alleles were missing from the IMGT database, and were also not present in VBASE2 (22) or the UNSWIg human heavy chain repertoire (23). Excluded from Table 2 are six alleles that were sequencing artifacts (further discussed in SI Text), and which were eventually removed by the genotyping process described later.…”

“…However, recent studies have discovered the presence of numerous V segments and alleles not reported in any published databases (6,(19)(20)(21), as well as a several novel D and J alleles (19). Some of these V alleles have been incorporated into the IMGT database or alternative databases of germline alleles [such as VBASE2 (22) or the UNSWIg human heavy chain repertoire (23)]. The completeness of the germline V(D)J database may greatly influence downstream analysis results, including clinically relevant decision processes (24), as unreported alleles can skew estimated segment distributions and because novel polymorphisms will appear as recurrent somatic mutations.…”

Sterile DJH Rearrangements Reveal that Distance Between Gene Segments on the Human Ig H Chain Locus Influences Their Ability To Rearrange

“…V H genes positioned closer to the D locus rearrange more frequently than V H genes positioned more distantly (9)(10)(11)(12). In humans, it is unknown whether this position bias also applies during the generation of the adult BCR repertoire.…”

Models and methods for analysis of lymphocyte repertoire generation, development, selection and evolution

“…The tools based on the IMGT databases, however, are subject to the problems which the Collins group has pointed out [55][56][57]. The next step is to assign the sequences into clonally related groups.…”