This inference was discussed during IARC Meeting 112 on Jan 9th, 2023.

TRBV12-4*01_C87T has been inferred in nine genotypes in the VDJbase P4 data set, including in VDJbase P4_I24_S1, a haplotypable data set (based on heterozygocity in TRBJ1-6). The genotype is also implied to carry TRBV12-4*01. No other gene apart from TRBV12-3 in the IMGT database is closely related to these alleles of TRBV12-4. The novel allele is the most expressed allele in the repertoire (72% allelic frequency; 1.47% of the total error-free population). It is represented by 499 error-free sequences and 467 unique CDR3s in the error-free set. Haplotyping based on allelic diversity in TRBJ1-6 demonstrates perfect separation from TRBV12-4*01. The data defining the 3’-end of sequences is negatively affected by trimming during the rearrangement process. IARC affirms the sequence (as TRBV12-4*i01) at Level 1 up to and including base 323 based on inference alone. It is acknowledged that the allele most likely carries 3 additional bases, typically AGC, at base positions 324-326. Analysis of sequences that have not been trimmed upstream of each position strongly suggest that the 3’-end of the sequence is indeed AGC. For many applications, IARC recommends the use of germline sequences representing their most likely full length base sequence. There is, however, insufficient evidence to make an affirmation of the full length sequence based on inference alone. In this case genomic data confirming the full length sequence is available from two other subjects (Rodrigues et al. Cell Genomics 2, 12, 100228 (https://doi.org/10.1016/j.xgen.2022.100228 (Supplementary Table S4)). Hence the full length sequence is approved based on inference in combination with genomic data. The locus that carries human TRBV genes is highly complex. Genes may be duplicated or deleted, and identical sequences may be found in more than one gene. The name (with an “i” allele designation) of an inferred allele does not imply that its precise genetic location is known. It just relates to the most similar allele presently found in the IMGT database, or to the gene with the lowest alphanumeric value, should alleles of multiple genes be equally matched to the novel allele in question. The only other similar gene has been mentioned above (TRBV12-3).