These sixteen RAW files represent twelve different samples submitted to the Centre for Proteomic and Genomic Research in Cape Town, South Africa, for shotgun proteomics analysis. Two different strains of Mycobacterium tuberculosis are included: --K636, a "Beijing" variant susceptible to TB drugs that was isolated in Worcester, Western Cape, South Africa --SAWC1129, an M.tb variant that is resistant to isoniazid, isolated in the Western Cape, South Africa Each of these two strains were split to two cohorts, one subjected to sulfamethoxazole (SMX) and another that was not. The total protein lysates were digested via FASP and then subjected to LC-MS/MS without prior fractionation. CPGR employed a Q-Exactive mass spectrometer on each sample, producing RAW files that span 130 minutes for each experiment. The fragmentation method was "HCD" at a collision energy of 26, with high resolution measurements in both MS and MS/MS scans. CPGR started with a test run of sample 9 (see "_01" on file name) but then ran each of the sample vials once afterward (it seems likely the initial 9 run was to test that methods were working appropriately). After each of the vials had been run, CPGR opted to re-run sample tubes 2, 3, and 4 to produce the files with 'a' suffixed to the tube numbers. The sample numbers may be understood by this legend: S1-S3 SAWC1129 SMX cohort (three replicates) S4-S6 SAWC1129 Control cohort (three replicates) S7-S9 K636 SMX cohort (three replicates) S10-S12 K636 Control cohort (three replicates) Only the K636 data have been described in a manuscript, at this point. We decided, however, to upload all the RAW data since they were produced as a single set. The quality metrics were produced by the "IDFree" mode of QuaMeter, available as part of the "Bumbershoot" package at ProteoWizard. The metrics, supplied in the "metrics.tsv" file, were produced using the configuration file provided as quameter.cfg. ProteoWizard msConvert was used to produce centroided mzML files from the RAWs, using the "zip" option. These spectra were identified in MS-GF+ 20170113 on a Linux machine using the sequence database "20170518-Ensembl-Mycobacterium_tuberculosis_h37rv-Cntms.ASM19595v2.pep.all.fasta." The command-line options for running MS-GF+ are given in msgf.sh. The mzID files were imported by IDPicker 3.1.9729, then filtered by requiring a PSM FDR of 0.01 and a minimum of three spectra overall per protein group. An empirical protein FDR of 3.43% resulted, spanning 1559 discernible protein groups. Individual RAW files were organized to an experimental hierarchy defined in Assemble.txt. Identification results were exported in a spectral count per protein per experiment pivot table (20180326-IDPicker-Spectral-Counts-Per-Protein) and a spectral count per experiment table (20180326-IDPicker-Spectral-Count-Summary). These identifications are different than the set featured in the corresponding publication; they are intended only to be illustrative of the sample content.