Human TCRs and CDR3s sequenced from healthy volunteers and HIV-infected patients, before and after 14 weeks of therapy

Having amplified and sequenced the TCR repertoires of these 100 samples, we identified their VJ recombinations using a modified version of Decombinator (called vDCR), TCR analysis software designed in our lab. We then made use of random barcode sequences introduced before amplification to error- and frequency-correct our Decombinator assignations (DCRs), before translating them and extracting their complementarity determining region 3 (CDR3) sequences.

Here we present the results of these analyses. The .dcrcdr3 files in this fileset consist of a unique DCR assignation per line, the CDR3 sequence it encodes, and the frequency with which that TCR appeared in the data following error-correction. Each line follows the format: 'V, J, Vdel, Jdel, insert: CDR3, freq' (V = V gene, J = J gene, Vdel = number of deletions from V, Jdel = number of deletions from J, insert = string of nucleotides from the end of the deleted V to the start of the deleted J, CDR3 = translation CDR3 sequence, from the second conserved cysteine residue in the V to the conserved phenylalanine of the FGXG motif of the J, freq = error-corrected frequency of that assignation.) The raw sequence data fastq files from which these TCRs were extracted is available in the Sequence Read Archive (SRA) under the Study accession number SRP045430.

The AccessionKey.xls spreadsheet cross-references all filenames with their appropriate SRA individual accession numbers. The Python scripts which were used to generate this data from that raw fastq data are also available on figshare (see links below).