- Fixed a bug that crashes when C-term mod mass is below -57Da.
- Added a test that checks whether the output path is valid before processing the data.
- Optimization for multithreaded performance.
- Several minor bug fixes.
- New scoring parameters are added for HCD/Q-Exactive/Trypsin/TMT. Parameters for HCD/HighRes/Trypsin/TMT have also been changed. As a result, for HCD spectra of TMT peptides, the number of identifications has been significantly increased.
- MS-GF+ now automatically recognize "TMT6plex" in the modification file and change the protocol to TMT.
- For mzML files converted from Thermo raw data using msconvert, MS-GF+ now reads monoisotopic precursor m/z first from a user parameter ([Thermo Trailer Extra]Monoisotopic M/Z instead of the CvParam "MS:1000744".
- The maximum number of variable modifications (per peptide) is written to the mzIdentML output file.
- Previously MS-GF+ ignored spectra having less than 20 peaks for non-TOF spectra. Now it ignores spectra having less than 10 peaks.
- Change in database indexing format. The index file keeps non-standard amino acids (characters other than 20 standard residue characters). Previously, non-standard amino acids were converted into ‘?’ while indexing. It caused a problem when converting mzid into pepXML using idconvert (ProteoWizard).
- Fixed a bug that creates duplicate PeptideEvidence items when more than one matches are reported for a spectrum (e.g. -n 10),
- Fixed a bug that crashes MS-GF+ when a modification makes the amino acid mass 0 or smaller (e.g. M-131).
- Added parameters for CID-LowRes-NoCleavage.
- Misc bug fixes.
- Previously separate SpectrumIdentificationItems were created for the same peptide if "pre" is different (e.g. R.SIPDSMNYGDEEENK and K.SIPDSMNYGDEEENK). Now they show up as the same SpectrumIdentificationItem. If the score is different due to different NTT (e.g. G.SIPDSMNYGDEEENK), a separate SpectrumIdentificationItem is created.
- Bug fix: falling into an infinite loop while reading an mzML file if it contains <binaryDataArray> of encodedLength="0"
- Previously, spectra are ignored in the search if the number of peaks is less than 20 for all types spectra. Now, for TOF spectra (i.e. -inst 2), this number has been changed to 3.
- The following features are added in the SpectrumIdentificationItem when "-addFeatures 1"
- MeanErrorAll: Mean of mass errors of explained fragment ion peaks
- StdevErrorAll: Standard deviation of mass errors of explained fragment ion peaks
- MeanErrorTop7: Mean of mass errors of 7 fragment ion peaks with highest intensities
- StdevErrorTop7: Standard deviation of mass errors of 7 fragment ion peaks with highest intensities
- Added TMT scoring model for HCD/HighRes (-m 3 -inst 1)
- Bug fix: duplicate PeptideEvidence id in mzid output
- Bug fix: crashing while using user-trained scoring parameters
- Added Q-Exactive unlabeled phosphorylation parameters
- Bug fix: recognizing ETD+SA spectra in mzML as CID
- Minor bug fix: mistakenly assigning 3 mods to a N-term amino acid.
- Added scoring parameter sets for Q-Exactive iTRAQ (-inst 3 -protocol 2) and iTRAQ phosphopeptide enriched (-inst 3 -protocol 3) samples.
- When "-protocol" parameter is missing, MS-GF+ automatically selects an appropriate protocol depending on the modification file.
- In the mzid output, Enzyme has attributes "missedCleavages".
- Scoring parameters for Q-Exactive (-inst 3) have been added.
- A bug (crash with an exception) in edu.ucsd.msjava.ui.ScoringParamGen has been fixed.
- "-ti" parameter accepts only two comma separated integers, i.e., "-ti 1" will be rejected.
- MzIDToTsv adds "Title" column to the tsv file if the input spectrum is the mgf format.
- "No enzyme" (-e 1) was renamed to "unspecific cleavage".
- "No enzyme (peptidomics)" (-e 9) was renamed to "no cleavage".
- <Enzyme>...</Enzyme> in the mzid output now shows right CVs for -e 1 and -e 9.
- The following bug has been fixed
- Precursor errors are negated when MzIDToTsv was run.
- C-term specific modifications do not show up in result files.
- The following bug has been fixed
- Ignoring some spectra when -n value is larger than 1
- The following bugs have been fixed
- Profile spectra should be ignored but not.
- Peptides shorter than MinLength are reported.
- Redundant proteins are reported in the same line in tsv files
- Some SpectrumIdentificationResult elements have no SpectrumIdentificationItem element.
- Different PeptideEvident elements have the same ID.
- Bug fix: crashing while reading mzML files converted from wiff
- Fixed a bug of ignoring N-term peptide when N-term Met cleaved peptide exists.
- Add scan numbers for NIST mgf files
- Fixed a bug in edu.ucsd.msjava.misc.MS2ToMgf
- Fixed bugs to ignore N-term fixed mods in the output
- Added edu.ucsd.msjava.misc.MS2ToMgf
- Fixed bugs reporting NaN as additional features when spectrum has charge 0
- Updates to handle multiple charge states in the ms2 format
- Fixed the bug to erroneously report precursorMz in converted tsv files
- Update MS2 parser to support http://noble.gs.washington.edu/proj/crux/ms2-format.html
- Old MS2 format (http://fields.scripps.edu/sequest/SQTFormat.html) is no longer supported
- Fix the following bugs:
- Same PeptideEvidence occurs more than once
- SearchModification/SpecificityRules is always empty
- Unknown modification cvParam does not have cvRef
- SpectrumIdentificationResult has no SpectrumIdentificationItem
- MzIDToTsv -o TSVFile is now an optional parameter.
- The extension of target/decoy concatenated database file has changed from .revConcat.fasta to .revCat.fasta.
- *.revConcat.fasta: decoy proteins have prefix "REV_" (used by MS-GFDB)
- *.revCat.fasta: decoy proteins have prefix "XXX_" (used by MS-GF+)
- Bug fix: N-term or C-term residue-specific modifications (e.g. pyro-glu from Q) will have locations 1 (N-term) or length (C-term).
- Fix a bug to write SpectrumIdentificationResult twice for some spectra without precursor charges
- Fix a bug in parsing Agilent mzML files
- Fix the bug ignoring PSMs with IsotopeError=0, when MinIsotopeError was negative
- Change the '-ntt' parameter to set the minimum number of termini following the enzyme specificity rule
- Change the default value of '-ntt' from 1 to 2
- Fix the bug to set "fixedMod=true" for variable modifications
- Fix the bug not showing fixed modifications in the peptide list
- Fix the bug to show the location of N-term modification as 1 instead of 0
- Fix the bug to calculate massToChargeRatio as neutral masses instead of charged masses
- Replace the cvParam showing the dissociation method with a userParam named AssumedDissociationMethod in the SpectrumIdentificationItem
- Provide MzID to TSV converter (edu.ucsd.msjava.ui.MzIDToTsv)
- Ignore profile spectra in the analysis
- Support JRE 1.6
- Fix minor bugs
- Add the scan number CV in the output
- Fix minor bugs