Workflow Mascot Distiller Mascot IRMa

Workflow 3: Mascot Distiller/Mascot/IRMa-hEIDI/Spectral Counting. Data were processed automatically using Mascot Distiller software (version 2.4.3.0, Matrix Science). ESI-TRAP was chosen as the instrument, trypsin/P as the enzyme and 2 missed cleavages were allowed. Precursor and fragment mass error tolerances were set at 5ppm and 0.8Da, respectively. Peptide variable modifications allowed during the search were: acetylation (Protein N-ter), oxidation (M), whereas carbamidomethyl (C) was set as fixed modification. The IRMa software v1.31 [11] was used to filter the results. Filters used were: (1) peptides whose score≥query homology threshold (p<0.5) and rank≤1 are marked as significant; (2) Single match per query filter was: Move to ambiguous all peptides which are not assigned to best protein for this query (best is higher protein score); (3) FDR seeker filter: seek a 1% FDR based on score filtering; (4) Accession filter: Delete proteins coming from reverse database; (5) Specific peptide filter: accept only protein hits whose specific peptides count >=1. The filtered results were then compiled and structured within dedicated relational Databases and a homemade tool (hEIDI) was used for the compilation, grouping and comparison of the proteins from the different samples, analytical replicates and conditions to compare (Hesse et al., in preparation). In such workflow, total spectral count values calculated for each protein groups are used for quantification. Workflow 4: ExtractMSn/Mascot/Scaffold/Spectral Counting. Peaklists generation and protein identifications were made as detailed in workflow 1. Mascot results were loaded into the Scaffold software (Version 3.6.5, Proteome Software, Portland, USA). To minimize false positive identifications, results were subjected to very stringent filtering criteria as follows. For the identification of proteins, a Mascot 5-lipoxygenase score had to be minimum 30 and above the 95% Mascot significance threshold (\"Identity score\"). The target-decoy database search allowed us to control and estimate the false positive identification rate of our study, and the final catalog of proteins presented an estimated false discovery rate (FDR) below 5%. The spectral count metric used for quantitation corresponds to the Unweighted Spectrum Count values in Scaffold. Workflow 5: ExtractMSn/Mascot/MFPaQ/MS Signal analysis. The first steps (peaklist creation, database search, validation) were the same than in workflow 1. Quantification of proteins was then performed using the label-free module implemented in the MFPaQ v4.0.0 software, as previously described [3,12]. Briefly, the software uses the validated identification results and retrieves the XIC of the identified peptide ions in the corresponding raw nanoLC-MS files, based on their experimentally measured RT and monoisotopic m/z values. Peptide ions identified in all the samples to be compared are used to build a retention time matrix and re-align in time LC-MS runs. For peptides not identified by MS/MS in a particular run, this re-alignment matrix is used to perform cross-assignment and extract their XIC signal starting from a predicted RT. Normalization across conditions is performed based on the median of XIC area ratios for all the extracted peptide ions. Protein quantification is based on a protein abundance index calculated as the average of XIC area values for at most three intense reference tryptic peptides per protein. Workflow 6 and 7: Andromeda/MaxQuant/MS Signal analysis. The first steps (database search with Andromeda and validation) were the same as in workflow 2. For quantification purposes, either Intensities (workflow 6) or LFQ [13] (workflow 7) calculated by MaxQuant were used. The LFQ metric, as described in [13], is derived from the raw intensities by the MaxLFQ algorithm, which uses a specific normalization procedure, as well as a particular aggregation method to calculate protein intensities, by taking into account, for each protein, all the peptide ratios measured in all pairwise comparisons of the different quantified samples. “Match between run” time window was set to 2min. For LFQ quantification, only protein ratios calculated from at least two unique peptides ratios (min LFQ ratio count=2) were considered for calculation of the LFQ protein intensity.