Changelog

All notable changes to this project are documented in this file.

4.1 - 2024-05-29

• Require Java 11+
• DDA+ mode supports ddaPASEF data
• Significantly faster loading of Thermo raw files
• Calculate and report more modification-related scores for PSMs identified in open and mass-offset search workflows (especially useful for chemoproteomics workflows). The following columns added to PSM.tsv files:
  • ptm_mass: mass of the modification
  • localization_peptide: a peptide string with lowercase amino acids to indicate the sites with the best score
  • best_positions: a list of sites with the best score in the format <amino acid 1><site index 1>:<amino acid 1><site index 1>:.... For example T(4)I(5).
  • number_best_positions: number of the sites with the best score
  • score_all_unshifted: the score when matching only unshifted fragments (unshifted fragment is a fragment not containing any modification mass)
  • score_best_position: the score when putting the modification on the best (i.e. higest scoring) site
  • score_second_best_position: the score when putting the modification on the second best site
  • score_shifted_all_positions: the score computed based on all possible shifted fragments (useful for identifying PSMs labeled with probes that can bind to multiple sites on the peptide, such as in some PAL data)
  • position_scores: a string with peptide sequence and scores when putting the modification on the specific site. For example P(2.32)E(2.41)P(2.57)T(5.36)I(5.36)D(3.87)E(2.56)K(1.34).
  • ions_all_unshifted: the number of matched fragments when matching only unshifted (unmodified) fragments
  • ions_best_position: the number of matched fragments when putting the modification on the best site
  • ions_second_best_position: the number of matched fragments when putting the modification on the second best site
  • ions_shifted_all_positions: the number of matched fragments when putting the modification on all sites and matching all possible shifted (modification-containing) fragments
  • score_shifted_best_position: the score when putting the modification on the best site and matching only the shifted fragments
  • ions_shifted_best_position: the number of fragments when putting the modification on the best site and matching only the shifted fragments
  • shifted_only_position_scores: similar to position_scores but only matching shifted fragments
  • shifted_only_position_ions: similar to shifted_only_position_scores but showing the number of matched shifted fragments
• Add analyzer_types parameter
• Add a new mass calibration and parameter optimization mode, 4, for calibrating the mass and adjusting only the fragment mass tolerance
• Propagate instrument info to calibrated and uncalibrated mzML files
• Various bug fixes and improvements

4.0 - 2023-12-19

• Implement new MSFragger-DDA+ mode, triggered when DDA data is annotated as DDA+ (MSFragger parameter data_type=3). MSFragger will perform full isolation window search, identifying multiple co-fragmented peptides from chimeric DDA spectra. This mode significantly boosts the number of IDs compared to conventional DDA search.
• Implement detailed mass-offset search mode. Each specified mass offset can have their own modification sites, neutral losses, and diagnostic ions.
• Overhaul the mass calibration and parameter optimization procedure for DIA data
• Significantly improve the ddaPASEF .d loading speed
• In the pin file, write all proteins in one column
• Extract and print isolation window sizes when checking the LC-MS files
• Replace ‘peptide_length’ score with Boolean peptide length scores (length < 8, 8, [9, 30], and > 30) in pin files
• Remove log10_kl_negative_1 and log10_kl_negative_2 from the DDA pin file
• Remove add_topN_complementary option
• Various bug fixes and improvements

3.8 - 2023-05-31

• Make the decoy prefix support #
• Add group FDR variable support
• Add require_precursor option for DIA data
• Add reuse_dia_fragment_peaks option for DIA data
• Remove the output_report_topN constraint for the DIA data
• For DIA data, do not remove precursor peaks when the isolation window larger than 20 Th
• Set override_charge to 0 for the first search
• Do not automatically set remove_precursor_peak to 0 for DIA data
• Various bug fixes and improvements

3.7 - 2023-01-12

• Add write_uncalibrated_mgf parameter that only works with .raw and .d formats
• Fix a bug reporting the incorrect prevAA and nextAA due to the clipped Methionine
• Various minor bug fixes and improvements

3.6 - 2022-12-03

• Require Java 9+
• Generate _(un)calibrated.mzML files. Stop generating _(un)calibrated.mgf files
• Various minor bug fixes and improvements

3.5 - 2022-05-27

• Require Java 9+
• Add remainder_fragment_masses, activation_types, and min_sequence_matches parameters
• Do not distinguish I and L
• Write ion mobility and compensation voltage to <run name>.tsv file.
• Add modification score to the pin file from open and mass-offset searches
• In pepXML file, change native_id to SpectrumNativeId to follow the pepXML schema
• For Thermo raw/mzML/mzXML data, using the native scan number as the scan number
• Clean up the protein description according to ProteinProphet’s rule
• Print all decimal points in the pepxml files
• Various minor bug fixes and improvements

3.4 - 2021-10-30

• Support N-terminal enzymes
• Support two enzymes
• Adjust scores in pin files
• Automatically adjust use_topN_peaks, minimum_ratio, output_report_topN, remove_precursor_peak, intensity_transform and precursor tolerances for DIA mode
• Check data type by isolation windows width
• Add 450 and 500 top-N peaks to the parameter optimization for low-resolution data
• Optimize max_fragment_charge for low-resolution data
• Always write mass offsets to pepXML header for mass offset searches (to enable IonQuant after PTM-S)
• Various minor bug fixes and improvements

3.3 - 2021-07-22

• Support prmPASEF in Windows platform.
• Add use_all_mods_in_first_search parameter to use all specified variable modifications in first search.
• If users specify variable modifications on cysteine, use them in first search.
• Add retentiontime, isotope_error, and mass_diff_score columns to pin files.
• Remove abs_mass_diff, matched_ion_fraction, delta_hyperscore, and abs_ppm columns from pin files.
• Add protein_descr and modified_peptide attributes to pepXML file.
• Always write calibrated.mgf file for .raw and .d files.
• Do not write tsv file by default.
• Improve peak denoising for low-resolution MS2.
• Calibrating mass by median shift if there are only more than 100 high quality PSMs.
• Improve the parameter optimization step.
• Various minor bug fixes and improvements.

3.2 - 2021-03-09

• Add FAIMS support.
• Tracing fragment and refinement of scores in the DIA mode.
• Support directly searching DIA data without any MS1 scans in the DIA mode.
• Support fragment mass calibration in the DIA mode.
• Add data_type = 0/1/2 option to specify data types: 0 = DDA (default), 1 = DIA, 2 = DIA with narrow windows (e.g. GPF data).
• Support removal of neutral loss fragment peaks (default), and add deneutralloss = 0/1 option.
• Check MS2 resolution before the search.
• Require at least two sequence ions matched in the Glyco mode.
• Add check_spectral_files = 0/1 option to turn on and off file checking.
• Always write _calibrated/_uncalibrated.mgf for ddaPASEF data (.d).
• Filter out scans outside of [200, 15000] range.
• Update pepXML schema to version 1.2.2.
• Improve help doc and change --config to --config closed/open/nonspecific/glyco.
• Remove output_file_extension option.
• Various minor bug fixes and improvements.

3.1.1 - 2020-10-01

• Fix a bug that does not write calibrated MGF.
• Fix a bug that do not allow labile modifications in mass offset search with restrict_deltamass_to = all.

3.1 - 2020-09-30

• New DIA search mode of MSFragger, supporting searching DIA data with parameter precursor_mass_units = 2/3.
• Precursor isotope error correction with parameter precursor_mass_mode = corrected.
• Support restricting delta mass (open/mass offset searches) to certain amino acids with parameter restrict_deltamass_to.
• Support putting localized delta mass to peptides as variable modifications with parameter mass_diff_to_variable_mod = 0/1/2. For N-Glyco data, will also put unlocalized glycans to the “N” of the first “N-X-S/T” sequon.
• Support pin format output, which is compatible with Percolator.
• Upgrade timsdata library (for Bruker timsTOF data) to 2.7.0.
• Write localization information to pepXML file with tag ptm_result.
• Added b~ and y~ fragment ions to localization (Glyco mode)
• Reduce the PSMs required for mass calibration from 500 to 250, and increase the expectation threshold from 0.001 to 0.005.
• Improve parameter optimization procedure.
• Add a MS2 resolution check.
• Change “O”’s mass to pyrrolysine’s mass, and print two comments for “O” and “U”.
• Change the default value of allowed_missed_cleavage to 2.
• Change the default enzyme to stricttrypsin.
• Print more significant digits for masses and other key parameters in the output file.
• Various minor bug fixes and improvements.

3.0 - 2020-06-05

• New Glyco/Labile mode of MSFragger (for N-linked and O-linked glycopeptides; can also be used for other labile modifications).
• Change the default value of write_calibrated_mgf to 0.
• Add 7 ppm to nonspecific search’s fragment tolerance list (optimization step).
• Various minor bug fixes and improvements.

2.4 - 2020-03-21

• Write original spectral file extension to pepXML after mass calibration.
• Write native_id to pepXML give mzML or mzXML file.
• Add 5 ppm and 50 ppm to the fragment mass tolerance list.
• Improve printed message.
• Support non-zero mass offsets in first search, mass calibration, and parameter optimization.
• Various minor bug fixes and improvements.

2.3 - 2020-01-30

• Add a MS/MS deisotoping module, with a new input file parameter deisotope. Default is set to 1 (use deisotoping).
• Support water-loss ions: ion types b-18 and y-18.
• Add one more column to variable modification setting specifying the maximum allowed number of instances of that particular modification on the peptide.
• Replace max_variable_mods_per_mod with max_variable_mods_per_peptide that specifies the maximum allowed number of variable
  modifications on the peptide.
• Print the number of peptide candidates with the number of modifications exceeding max_variable_mods_combinations.
• Print all decimal points to the tsv file.
• Change the minimum allowed precursor_mass_lower to -230 (in open search).
• Improvements in the parameter optimization procedure.
• Increase estimated memory usage by including the sizes of spectra and result.
• Various minor bug fixes and improvements.

2.2 - 2019-11-08

• Check spectral files at the beginning.
• Add a parameter (write_calibrated_mgf) to write calibrated spectra to MGF files.
• Add uncalibrated_precursor_neutral_mass attribute to pepXML file.
• Change the default value of allow_multiple_variable_mods_on_residue to 0.
• Change the default values of min_fragments_modelling in open search to 2.
• Change the default values of min_matched_fragments in open search to 4.
• Limit the minimum value of precursor_mass_lower to -150 Da.
• Check intensity_transform and remove_precursor_range in parameter optimization step.
• Improve parameter optimization algorithm.
• Various minor bug fixes and improvements.

2.1 - 2019-09-08

• Fixed bugs.
• In parameter optimization step, improve the top-N peak candidates.

2.0 - 2019-09-05

• Switching from ‘date’ to ‘version number’ in naming released versions. This version is named 2.0 in recognition of the many improvements implemented in MSFragger in the last six months.
• Support direct reading from Bruker raw files (.d folder).
• Support MGF file converted by Bruker DataAnalysis.
• FragPipe now supports running MSFragger with raw Thermo or Bruker files as input within FragPipe (limited functionality).
• Pass ion mobility information (if applicable) from the spectral file (mzML, MGF, and .d folder) to the resulting pepXML file.
• New option to remove unfragmented precursor peaks from tandem mass spectra (remove_precursor_peak and remove_precursor_range parameters). Recommended option for ETD/EThcD data.
• New option to apply sqrt root transform to fragment ion intensities (intensity_transform parameter). Recommended option for ETD/EThcD data.
• Support calibrating precursor masses even when there are no MS1 scans in the input spectral file.
• Support assigning mass offsets to the localized amino acid sites.
• Add a --version flag to print version to console.
• Add a --help flag to print help information to console.
• Add a --config flag to write three fragger.params templates to current folder.
• Limit maximum precursor charge to 7.
• Various minor bug fixes and improvements.

20190628 - 2019-06-28

• Support direct reading from Thermo RAW files through the command line and the ProteomeDiscoverer node. FragPipe will be able to read raw files soon.
• Add an excluded_scan_list_file parameter that takes the path of a text file containing scan names. MSFragger would skip those scans if the path is not empty. Comment or delete this parameter name and its value if you do not want to use it.
• If there is no spectral file specified, generating pepindex and exiting.

20190530 - 2019-05-30

• Support mass calibration.
• Support optimal parameters finding.
• Support additional ion series including a-, b-, c-, x-, y-, and z-ions.
• Write localized modifications to .tsv file if localize_delta_mass = 1.
• Improve .tsv file’s columns.
• Support command line parameters.
• Print parameters to the command line.
• Write parameters to the pepXML file.
• Re-implement alternative protein mapping for efficiency.
• Make the MGF file parsing more robust.
• Fix a bug about the inaccuracy of the searching speed printed to the command line.
• Showing the numbers of unique peptides in the command line.
• Various bug fixes and improvement.

20190222 - 2019-02-22

• Improved error messages for OutOfMemoryError and when too many peptides were generated durning digestion
• report_alternative_proteins option added. Turned off by default.
• write input file type in msms_pipeline_analysis tag, previously was hardcoded as ".mzXML"
• write alternative proteins
• If MsAdjuster is used, output precursor intensity in spectrum_query tag
• added option to write both TSV and pepXML files. output_format = tsv_pepXML

• added precursor_mass_mode option

  usage: precursor_mass_mode = isolated / selected / recalculated

• default shifted_ions_exclude_ranges to (-1.5, 3.5)

20181110 - 2018-11-10

• update MSFTBX dependency

20180924 - 2018-09-24

• added shifted_ions_exclude_ranges option

  example usage: shifted_ions_exclude_ranges = (-1.1033548378, -0.9033548378), (-0.1, 0.1), (0.9033548378, 1.1033548378), (1.9067096756, 2.1067096756), (2.9100645134, 3.1100645134), (15.894915, 16.094915), (42.905814, 43.105814)

• added output_location option

20180924 - 2018-09-24

• \n is used as a newline for all platforms
• when in split search mode, output files in the working directory instead of the directory of input files

20180912 - 2018-09-12

• fragment matching is based on neutral masses, memory usage reduced
• restrictions on digested peptide size relaxed
• read in Corrector input if present
• directory for temporary files can be specified with temp_location
• non zero exit codes when terminating with error

20180316 - 2018-03-16

• Fix display issue with static terminal modifications.
• Open search times are reduced by about a quarter.
• Fixed bug with peptide mass calculation less termini.

20180224-RC1 - 2018-02-25

• MSFTBX upgraded to version 1.8.2.

20180223-RC1 - 2018-02-23 (public release)

• Java 9 compatible JAR
• Protein names are escaped to produce valid pepXML

20180215 - 2018-02-15

• Added support for mass offsets (enabling multi-notch searches)