Releases: smith-chem-wisc/MetaMorpheus
Releases · smith-chem-wisc/MetaMorpheus
0.0.306
- Added a beta version of a glycopeptide search task
- New MetaMorpheus releases are automatically uploaded to Docker Hub
- Added q-value filtering to mzID output
- Added options in the search task for how to label contaminants and targets with the same accession
- Fixed a crash in the GPTMD task when searching low-resolution CID data
- Fixed a bug where selecting "Yes and don't ask me again" when being asked if MetaMorpheus should be closed did not save the preference to not ask again
0.0.305
- VERY IMPORTANT: This version (and future versions) of MetaMorpheus requires that you install the .NET Core 3.1 Runtime from Microsoft.
Mac: https://dotnet.microsoft.com/download/dotnet-core/thank-you/runtime-3.1.3-macos-x64-installer
Linux: https://docs.microsoft.com/en-us/dotnet/core/install/linux-package-manager-ubuntu-1910
- MetaDraw can now display some basic statistics from search results (histogram of precursor PPM errors, hydrophobicity vs. retention time, etc.)
- The machine learning PEP algorithm is faster and trains on more features
- In the output options in the Search task, you can compress or turn off individual file results
- Fixed a crash in low-resolution searches (LowCID) when a scan has no peaks
0.0.304
- Upon selecting certain proteases in the calibration/GPTMD/search tasks, a popup appears which recommends settings and asks if they should be applied
- The classic search is a bit faster and more memory-efficient
- Modern search is now considerably faster for open-mass searches
- The crosslink search is faster and more accurate
- Fixed several bugs in the command-line version of MetaMorpheus for Linux/OSX (file paths were not passed in properly, making the program difficult to run)
- Fixed a bug in low-resolution CID/HCD searches that caused high memory usage
0.0.303
- The Percolator output has been updated to include more features to train on. This set of features is identical to the set of features that MetaMorpheus's machine-learning PEP calculation trains on
- You can now edit the default settings for MetaMorpheus's tasks. To save a task as default, create a new task, edit the settings you want, and then click the "Save As Default" button. You can access saved default .tomls by going to Data -> Open folder with mods/data files -> DefaultParameters. Be aware that updates to MetaMorpheus may add new settings; these new settings will not exist in old default .toml files. Any new setting that is not specified in a .toml will take the default value for that setting
0.0.302
- MetaMorpheus has a new icon!
- A machine-learning algorithm similar to Percolator is used to estimate PSM-level error probabilities (the posterior error probability, "PEP"). This replaces the e-value calculation
- MSFileReader is no longer needed to read Thermo .raw files. Just drag your files in and accept the Thermo licence! Reading .raw files is faster now too
- Added protein ambiguity to the crosslink search
- In the crosslink search, if a single peptide outscores the alpha peptide, the single peptide is preferred. This helps correctly assign spectra with loop-linked missed cleavages
- Added many UniMod accession numbers to MetaMorpheus's modifications, which are reported in the .mzID file
- Fixed a bug where the nativeID for each PSM in the .mzID file was not correctly formatted
- Fixed a bug where sometimes parsimony was very slow when many PTM-containing peptides were present and "Treat modified peptides as different peptides" was unchecked
- Fixed a bug where the text boxes to search for a modification in the GUI would stop working
0.0.301
- The Search task settings have been grouped into "basic" and "advanced" settings
- SILAC quantification has been added!
- Custom amino acids are now supported. Go to "Data -> Create new amino acid". Protein databases containing the specified letter will be read as the defined custom amino acid
- G-PTM-D now does not annotate a mod in the database if an existing UniProt mod with the same mass already exists on that residue
- Crosslinkers are now stored in a file so that custom crosslinkers can be stored without needing to re-make them every time
- Proteins removed by the "top-picked" FDR method are now re-added with a conservative estimate of their FDR
- Additional peak-filtering options have been added to the advanced settings menu in the Search task
- Fixed a bug in the "PSM Count (unambiguous, <0.01 q-value)" output column
- Fixed a few bugs in the fast Non-Specific search mode
- Fixed a crash if a spectra file has no MS2 scans
0.0.300
- Fixed a somewhat rare bug where proteins with ">" in the protein name had truncated names when read from a .fasta database. This also caused a crash in modern search. This bug was present in v0.0.298/299
- Fixed a bug where the peptides listed in AllPeptides.psmtsv did not have the correct value under "PSM Count". The header for this column has been changed/clarified to "PSM Count (unambiguous, <0.01 q-value)"
0.0.299
0.0.298
- MS3 and mixed-mode MS2 (e.g., HCD/ETD) searches are possible in the crosslink search now. This is a preliminary implementation, and these extra scans are not used for scoring yet, but you can search them and then view them with MetaDraw
- Added some MetaDraw features (font size, bold font, etc.). MetaDraw also has a more minimalistic look
- The warning for not having C++ redistributable installed is shown immediately on startup now instead of when adding a .raw file
- Fixed a bug where peptide-level q-values were not calculated if "write decoys" was unchecked in the search task
- Fixed a bug in naming protein isoforms
0.0.297
- Low-resolution HCD/CID data is now searchable with better results, using XCorr instead of the Morpheus score. To use it, change dissociation type to "LowCID"
- Fixed a crash in calibration if modifications without chemical formulas were used
- Fixed crashes related to different regions formatting decimal numbers differently
- Fixed a match-between-runs crash