Supplement to "Direct maximization of protein identifications from tandem mass spectra"

Mass spectrometry data

• Yeast spectra are available here. The data sets "Yeast 1," "Yeast 2" and "Yeast 3" were used in this work.
• The original worm spectra are not available, but the search results are stored in SQT format here: full set, randomly selected subset. The corresponding C. elegans database is here. In addition to worm proteins, the database contains 4311 E. coli proteins and 44 common contaminants. Each protein occurs in the database twice, once in the forward and once in the reverse orientation. The reversed proteins have IDs of the form "random_seq_1".
• Human spectra are available here. Each set of spectra is available in both MS2 and compressed MS2 format. Crux search-for-matches output is stored in SQT format. The human reference database used for the searches is here. In addition to human proteins, the database contains 47 common contaminants. Each protein occurs in forward and reverse form.

External gold standard

• mRNA transriptome data from F. C. P. Holstege, E. G. Gennings, J. J. Wyrick, T. I. Lee, C. J. Hengartner, M. R. Green, T. R. Golub, E. S. Lander, and R. A. Young. "Dissecting the regulatory circuitry of eukaryotic genome." Cell, 95:717–728, 1998.
• Protein tagging data from S. Ghaemmaghami, W. K. Huh, K. Bower, R. W. Howson, A. Belle, N. Dephoure, E. K. O'Shea, and J. S. Weissman. "Global analysis of protein exression in yeast." Nature, 425:737–741, 2003.
• The following thresholds were applied to the above data sets: 1) all 1060 proteins whose mRNA copy count was higher than the average copies/cell counts (6.10929) were considered present according to the microarray experiments, and 2) all 449 proteins detected by both GFP (green fluorescent protein) and TAP (a specific antigen) with intensity above average intensity (12066 and 0.54694, respectively) were considered present according to the protein tagging experiment. The intersection of these sets, consisting of 313 proteins, was used in the validation experiments. Note that these numbers differ slightly from the ones reported in the published paper. The source of this discrepancy is unclear.