Versions of PEAKS

Past & Present PEAKS Versions


  • Deep learning-boosted ID for PEAKS DB
  • Streamlined DIA workflow (Specral Library Search + directDB + de novo)
  • DeepNovo-based peptidome workflow for immunopeptidomics.
  • Glycan module – a structure-based solution to glycoproteomics
  • Support ZenoTOF
  • Other minor improvements and bug fixes

  • Support of Thermo FAIMS Data
  • New Text format spectral library
  • Faster Spectral library search
  • New Spectral library viewer
  • Improved spectral library search interface
  • Updated Agilent file reader
  • Protein FDR filter
  • Improved LFQ and SILAC algorithms to limit missing values
  • Other minor improvements and bug fixes

New Features

  • Spectral library support
  • QC tool for LC-MS
  • Enhanced de novo sequencing using deep learning
  • ID-directed Label-free quantification for large scale proteomics
  • Support DIA, SWATH and diaPASEF

Improvements & Bug-fixes

  • LFQ statistical chart of before and after ID-transfer (missing values)
  • Improved LFQ sample-level and group-level correlation graphs
  • Visualization identification rate compared between sampleExport of matched fragment ions from de novo sequencing results
  • Integration of the latest X!Tandem version in PEAKS’ InChorus search
  • Ability to modify #significant digits for protein and peptide abundance value export
  • Protein group FDR bugfix
  • Direct .wiff data support

New Features

  • Support Q-TOF and TOF for DIA analysis
  • Support mascot server 2.6.x
  • Software mass calibration for DB search
  • Support Uniref database with taxonomy filtering
  • Redesigned Multi-round search to merge results from the previous round
  • Index column and protein count info in protein table
  • Redesigned user interface for DB result peptide details
  • Optional ratio columns in LFQ peptide table
  • LFQ peptide correlation plot
  • Export ion mobility (invk0) information in MGF, and mzXML, mzIdentML
  • Export ‘Found By’ column in DB peptide, psm and protein-peptide export
  • Bruker data loading option in Profile/Line mode

Improvements & Bugfixes

  • Copy config from previous PEAKS version
  • Protein FDR calculation
  • mzIdentML export speed
  • Custom Label Q method creation
  • Filtering Compare Run results
  • LFQ Runtime speed and accuracy
  • ‘All denovo candidate’ export to honor ALC filter
  • DIA PTM Search

Resolution for Chimeric Spectra

  • Using the new MS1 feature-based identification approach, one MS2 scan can be associated with multiple peptide features within its acquisition window, enabling multiple peptide identifications per chimeric MS2 spectrum.

Spectral-Library-Free DIA Support

  • PEAKS X’s new feature-based identification approach enables full DIA analysis support without the need of a spectral library. DIA spectra are analyzed directly to determine the peptide sequences of peptide features in their acquisition window, allowing peptide identifications beyond the information from a specific spectral library. With the combined de novo and feature-based technique, PEAKS removes biases found in commonly utilized approaches.

De Novo Sequencing for DIA Data

  • PEAKS X introduces the first de novo sequencing solution for DIA data to provide non-biased results.

Improved Identification of Endogenous and Low Abundant Peptides

  • PEAKS X improves analysis for endogenous peptide identifications by allowing variability at the C-terminus of peptides in no enzyme searching. This combined with feature-based peptide identification increases sensitivity when analyzing endogenous peptide datasets.

Ion-Mobility Spectrometry (IMS-MS) support

  • IMS-MS data can be analyzed with PEAKS de novo, identification and quantification work flows and interactive data visualization tools allow data view projected on m/z-rt or m/z-1/k0 dimensions.

Additional Improvements

  • Feature-based analysis for timsTOF data
  • Support for Thermo Fisher Scientific’s Advanced Peak Determination (APD)
  • Purity correction for TMT/iTRAQ
  • Added TMT11-plex to built-in methods
  • Added “Digest Mode” for sample analysis to support “No Digestion” samples
  • Protein group FDR in identification result
  • Added 1/k0 range in features table

Two methods for abundance calculation:

  • Summed area under the curve of the peptide isotope envelope
  • Summed maximum intensity over the retention time profile of the peptide isotope envelope

Label Free Quantification

  • Increased LFQ accuracy and sensitivity by improved feature detection

Improved SILAC Quantification module

  • Improved accuracy and sensitivity by enhanced SILAC pair feature detection and alignment
  • Ratio-focused quantification (e.g. super-SILAC)
  • Multiple group comparison (e.g. time series quantification)
  • More statistics:
    • Paired t-test for single-group analysis
    • Welch’s ANOVA for multiple group comparison

Label Free Quantification

  • Top-three unique peptides selected for LFQ
  • Improved LFQ accuracy by enabling exclusion of 1) peptides with both modified and unmodified forms, and 2) redundant peptides
  • Peptide-level quantification provided in peptide table
  • Sample clustering and correlation views

PTM Profiling

  • Enhanced PTM profiling in Identification result: multi-sample PTM comparisons; MS2 spectra comparison of modified v.s. unmodified peptides
  • Enabled PTM profiling in quantification (Q) result: confident PTM quantitation filtered by A-score or minimum ion intensity threshold; normalized by protein expression levels; PTM site clustering; modified peptide / feature details

Global Comparative View

  • Protein/peptide spectral counts and MS1 area in each sample provided in identification result for semi-quantitative analysis

EThcD Data Analysis

  • Advantages of EThcD fragmentation enables confident peptide identification

New TMT/iTRAQ Quantification module

  • Both MS2 and MS3 quantification are supported
  • Cross experiment comparison is supported through spike channels
  • Automatic protein FDR estimation is supported
  • Support manually removing low quality spectrum for quantification.

New SILAC Quantification module

  • Feature based quantification module
  • Replicate analysis is supported
  • Interactive validation for feature details
  • Interactive volcano-plot for proteins
  • Arginine to Proline conversion is corrected automatically
  • Easy sample grouping

Contaminant database support for identification

  • Support searching target database together with a contaminant database
  • Contaminant proteins can be shown or filtered out

Enzyme specific display for protein coverage

  • Display protein sequence in different enzyme format

Other improvements

  • Improved feature detection
  • Improved feature and PSM association
  • Improved label free quantification
  • Spectrum view supports absolute intensity and fragment ion error display
  • Support for ProteoWizard MS converter
  • Support for Thermo MS filereader 3.1 SP2
  • Improved PTM profile exporting
  • Support exporting for fragment ions
  • Better support for mzIdentML formats
  • Better multi-user support

PTM Profiling

  • A tool to compare and quantify detected modifications sites on proteins. Determines the protein’s PTM changes by comparing the “area-under-the-curve” from the LC-MS of the modified peptide features to the associated, unmodified (native) peptide features.

Peptide Mapping

  • A powerful comparison tool, which provides an intensity map of all the peptides associated with any particular protein ID. The maps can be analyzed in terms of the retention time vs intensity, or the peptide mass vs intensity. PEAKS allows users to compare peptide maps of up to 3 different data files.

Peptide Feature Intensity

  • One button to fully analyze and quantitate peptides across all samples.

Automated Multi-Round Search

  • A tool to launch multiple rounds of a database search using de novo only spectra.
  • De novo only spectra are high quality spectra with high scoring de novo results that could not be identified in the initial database search.

More Friendly User Interface

  • Direct link from peptide view/de novo view to LC/MS view
  • Save preferred PTM list for PEAKS PTM search
  • PEAKS can run smoothly under different language settings
  • Spectrum viewer can run more smoothly
  • Report more error messages when fails to open a PEAKS project

Algorithm Improvement

  • Improved speed of inChorus search on large data set
  • Improved feature detection efficiency for Waters QTOF data

Vendor Support

  • Bundles Thermo MSFileReader 3.0 SP3 with PEAKS. MSFileReader can be installed during PEAKS installation. The Config Wizard can also install the MSFileReader
  • Waters RAW data can be loaded in PEAKS out of the box. No further configuration is required
  • Fixed an issue in Agilent data loading

Bugs Fixed

  • Export all results from project when SPIDER has zero score hits
  • Empty inChorus run without PEAKS DB
  • Fail to load MZML file when empty peak list exists in file

  • Added Mascot 2.5 support in inChorus
  • Better support on Waters raw data
  • Fixed an issue related to usage when using a different user account
  • Fixed an issue in mzIdentML exporting when some fractions had no results
  • Fixed an issue reading Bruker raw data
  • Fixed an issue that caused OMSSA search to sometimes fail after a Mascot search

  • Improved 3rd party exporting to Scaffold and resolved PTM issue
  • Fixed mzIdentML protein reporting issue
  • Fixed a regression bug affecting charge and mass correction on high charge Orbitrap data
  • Third party and pepXML export from inChorus results are now disabled. It was unintentionally opened in the initial PEAKS 7 release. In mzIdentML, there are currently no CV term in the standard for PEAKS inChorus results.
  • Unique peptide filter in Label Free Quantification results now correctly uses the unique peptides instead of the peptide features as the filter
  • Fixed an issue in Label Free Quantification p-value calculation. Users do not need to redo the analysis. Re-open the results will have the p-value displayed correctly
  • PEAKS now supports 64 bit CompassXtract libraries from Bruker
  • PEAKS now can read the scan number from MGF files generated by the new Compass tools
  • Better handling on MGF/PKL/DTA files with excessive white spaces
  • Better handling on importing Mascot results into PEAKS. Solved issues for some custom FASTA header formats
  • Resolved slowe Derby initialization issue.
  • PEAKS project converter: MS1 spectrum information is correctly converted from PEAKS 6 projects

Improvements to de novo Sequencing

  • Interactive Protein Coverage View
  • New Automatic de novo Result Validation
  • Improved de novo Residue Local Confidence Score
  • CID/HCD/ETD Complementary Fragmentation Support

New Label Free Quantification Module

  • Enhanced Feature Detection – Expectation Maximum (EM) algorithm
  • Significant improvement to -10LgP Score for Peptide Features and Proteins
  • New RT Alignment Algorithm
  • Normalization using TIC or selected protein(s)

Projects, Processing & Review

  • Improved speed for PEAKS DB, PEAKS PTM and SPIDER
  • Much improved data 2D/3D heatmap visualization and navigation
  • New project wizard to facilitate the creation of complex project with a large number of samples and fractions
  • Reduced false discovery on PEAKS PTM algorithm

Collaborate

  • Better support for Thermo and Agilent raw files on 64bit OS
  • Pepxml exported from PEAKS can be imported into Skyline
  • Better support on mzidentML (version 1.1) format
  • Support pepxml exported from PD 1.4

Algorithm

  • New nonspecific enzyme digestion support. Now one can allow nonspecific enzyme digestion at 0, 1, or both ends of a peptide. Having more nonspecific digestion ends will increase the search sensitivity at a reduced search speed.
  • Consider neutral loss of phosphorylation and sulfation PTMs in the scoring functions of PEAKS DB and PEAKS PTM
  • Newly support protein N-term PTMs
  • New -10lgP score and FDR control for PTM and SPIDER results
  • FDR control of the inChorus result. Different search engines results can now be combined according to a unified FDR for each search engine.
  • Improved precursor mass correction. The mass correction function in “data refine” step recognizes the real monoisotopic mass even if the instrument’s raw data reported the isotope mass. The algorithm is improved in the new version.

GUI Improvements

  • Easier PTM selection interface. The PTMs are separated in three “common”, “uncommon”, “customized” to make it easy to specify the PTMs to search for. Additionally, a “Recent” list includes all the PTMs recently used by the user.
  • De novo only view added to inChorus result. The de novo only view reports peptides found exclusively by de novo sequencing. This useful view is now available to the inChorus result too.
  • Improved spectrum annotation view. Mouse over an amino acid in the spectrum annotation will highlight the supporting fragment ions. Setting an anchor peak will show mass different between the current peak and the anchor peak.
  • Improved project creation interface. The selection of instrument type and fragmentation mode is easier. Users can specify a proteolysis enzyme for each sample at the project creation interface.
  • More searching functions to locate a specific de novo sequence result. Now de novo sequencing results can be sorted and searched with the amino acid sequence.
  • Display absolute intensity in PEAKS Q. Users now can choose to display absolute intensities or ratios in PEAKS Q results. For ratio display, users can choose which sample the ratio is based.
  • More statistics in the summary view. The PTM profile table is improved. A protein FDR value is added.
  • One-click specification of common FDR values (e.g. 1%) in the FDR selection pane.

Better Community Support

  • A new configuration wizard to assist the download and installation of public protein databases and raw file readers/converters.
  • Major accuracy and sensitivity improvement on the analysis of AB SCIEX TripleTOF data.
  • Proteome Discover support. Now PEAKS can load the pepXML result file generated from Thermo’s Proteome Discover software.
  • Support mzIdentML result format. Now PEAKS can export mzIdentML file for downstream analysis such as Scaffold PTM.
  • Export high resolution images of the spectrum annotation and protein coverage view.
  • Export to website or single webpage format for easy sharing of results.

Unprecedented accuracy and sensitivity for database search

  • The revamped database search engine (PEAKS DB) results in substantially improved accuracy and sensitivity for peptide identification, causing the identification of more peptides with a reduced false discovery rate (FDR). In particular, the preview version of the PEAKS DB engine produced excellent performance in the ABRF/iPRG 2011 study for ETD data analysis.

Comprehensive result visualization

  • Numerous improvements were made to support the visual examination of the results from all different angles. To enhance user confidence, PEAKS provides tools for vigorous yet convenient result examination.
  • Users will also notice many areas where a summary of the result’s statistics are automatically provided. Through these charts users can easily answer important questions such as whether the target-decoy FDR calculation is reliable, and whether the instrument is well-calibrated. Additionally, the peptide and protein tables were rewritten to provide convenient search and sorting functions, allowing easy location and examination of certain particularly interesting peptides or proteins.

Built-in result validation

  • PEAKS DB provides a seamlessly built-in result validation system with an enhanced target decoy method. Score thresholds can be conveniently selected from the FDR curve, avoiding the guess work performed with other software.

Enhanced result reporting

  • The new summary view provides a central place for specifying score thresholds to filter results. Results can be filtered quickly and efficiently with the summary view showing the changes at a glance. Filtered results can easily be exported to several CSV and HTML files for publication and/or result sharing with non-PEAKS users. Furthermore, results can be saved as a PEAKS project to be viewed by your collaborators with a free PEAKS Viewer.

De novo sequence tag generation

  • By specifying a confidence threshold with a user-friendly sliding bar, users can promptly convert the acclaimed PEAKS de novo sequencing results into high confidence de novo tags. Users can even export these tags for integration with their own in-house analysis workflow.

Improved inChorus search

  • The inChorus function combines the search results of several search engines, including Mascot 2.3. An intuitive search engine icon displays which specific engines identified each peptide and when multiple engines agree. The new filtration rule allows users to filter the results flexibly, including by individual engine scores. A Venn diagram and side-by-side FDR curves display each engine’s contribution to the final result.

Heatmap view of quantification results

  • The heatmap provides a bird’s-eye view of the protein quantity changes across different groups of samples. The proteins are automatically clustered according to their quantity change patterns, facilitating the quick identification of the possibly interesting proteins and patterns.

Heatmap view for LC-MS data

  • The data heatmap provides a bird’s-eye view of the peaks in the LC-MS data, with peptide features highlighted.

New statistical peptide score

  • The scoring function for evaluating peptide-spectrum matches is significantly improved in 5.3. This is the foundation of the increased accuracy and sensitivity of PEAKS DB. The scoring schema of peptide identification involves matched peaks and their intensities, precursor mass error, enzyme specificity, de novo sequence, peptide size, and more. A statistical evaluation, -10lgP, is given for each peptide-spectrum match. Here lg() is the common logarithm with base 10, and P is the probability that a false identification of the current search has the same or better significance.

Algorithm

  • Overall improvement of software capacity, stability and speed
  • Increased de novo and database search accuracy for ETD data
  • Automatic precursor mass and charge correction for high resolution data
  • Improved the SPIDER scoring function
  • Reduced the memory overhead for SPIDER, improving its capacity and speed
  • PTM support in manual de novo
  • Improved workflows targeting different application needs such as Identification, Quantification and inChorus
  • Replicate analysis for label free quantification
  • FASTA database validation, important for custom databases
  • CID/HCD, CID/PQD support in iTRAQ quantification with Thermo instruments
  • Improved support for Thermo LTQ RAW file format
  • mzML/mzXML3.0/mzData/pepXML file format support

Analysis

  • MCP compliant result exporting
  • Result comparison interface
  • Improved spectrum display
  • Standardized PTM display
  • Optional display of assigned/unassigned de novo peptides in PEAKS Search results
  • inChorus Search parameter saving

  • enhanced project stability
  • enhanced search stability
  • PEAKS Q label free quantification
  • continued extremely large file support
  • homology seach: variable (and fixed) modification support
  • project converter tool (auto convert PEAKS .ANZ files to projects)
  • protein identification increased sensitive, less false positives
  • PTM Finder

  • improved protein identification and de novo sequencing precision
  • higher instrument calibration recognition
  • extremely large data file support
  • project based management approach
    • process multiple runs
    • compare, combine, filter the results together
  • decoy database searching
  • probability scoring
  • block search homology search mode
  • comparing and merging of CID and ETD fragmentation files
  • inChorus statistical charts and unified scoring
  • seamless connection with PEAKS Q

10% – 30% improvement in de novo sequencing accuracy on ion-trap data. Improvements to the inChorus search engine:

  • Better integration with Sequest and Mascot – searches can be launched from inChorus, and/or existing results imported
  • Existing results from any search engine can be loaded in (including PEAKS), avoiding having to repeat the search
  • A separate local database is defined for recompiling the results based on peptides from all search engines
  • The new RSD metric is available for all results, regardless of search engine

Improvements to instrument flexibility

  • Any combination of mass analyzer, fragmentation (incl. ETD) and ion source can be custom configured
  • For ease of use, PEAKS will automatically check the data to detect the scan mode

New SPIDER

  • Allows you to not only find homologous peptides, but also reconstruct the real sequence, confirming mutation and correcting de novo sequencing errors

Retention time prediction

  • RT is predicted for each peptide and you can filter search results by (realRT – predictedRT)

Data quality score

  • Allowing a second evaluation of sequencing results trustworthiness

Result Export

  • All parameters (start to finish) plus user’s notes included

Better Coverage

  • Improved charge and monoisotopic peak picking mean even better coverage

  • Post analysis filtering — results reports can be filtered by a number of criteria (for instance: score, delta mass, proteins with two good hits only, etc.)
  • A simple report to summarise de novo sequencing on a whole run, will be included. It will look like ‘peptide view’.
  • The scans may now be sorted by scan number, mass, m/z, or retention time
  • The de novo options screen allows users to limit the number of variable modifications allowed in a proposed de novo sequence.
  • Improved exporting:
    • Images of spectra, spectra with ions annotated, sequence-spectrum alignment, ion tables and error plots can now be exported in high resolution for printing. The usability of the export utility is improved.
    • WYSIWYG export of protein ID results, both from ‘peptide view’ and ‘protein view’ is enabled.
  • SPIDER speed has been improved several fold; its accuracy has also been improved.
  • Support for the HUPO organization’s proposed file format standard: mzData

  • SPIDER (Software Protein Identifier) is now available from a separate report in PEAKS. To assist in viewing the sequence tag matches and homology matches from the de novo sequences, a new addition to the search results report is presented — a new pane in the peptide view showing the alignment between the sequence found in the database and the de novo sequence.
  • Some errors have been corrected with regards to using the Data Refine tool on Thermo RAW data. RAW files that contain empty scans no longer cause troubles. Charge is recalculated from the spectral data where necessary, rather than relying on the scan header.
  • Memory usage enhancements mean larger datasets can be processed. PEAKS writes search results to disk to free memory, and is more explicit in its memory management during processing
  • Users can discover conserved regions in proteins, and easily distinguish between isoforms by way of a multiple sequence alignment tool. The new tool displays all matched peptides highlighted on the MSA. The MSA tool can be used on any number of proteins discovered after using PEAKS Protein ID, or inChorus protein ID.
  • A small error has been corrected in the assisted manual sequencing tools. Users performing manual sequencing with modifications will find it easier to complete a sequence. Additionally, the pop-up menu items have been re-named for clarity.
  • When configuring new databases, ‘sticky folders’ are used to make it easier to find the next file. As such, when you click the “browse…” button, the file chooser opens to the location of the last file you selected.
  • When exporting Protein results to an .xls file or .html file from the Protein ID report, we launch Microsoft Excel or the default web browser, respectively, and open the report.
  • The ion table can be configured by right clicking on it. One can now choose the number of decimal places to display, export, or switch between advanced and basic views.
  • Improved performance for Swiss-Prot database “initialization for the purpose of taxonomy based searching”.

  • Matrix Science: Mascot and Thermo Sequest results can be viewed right along side of the PEAKS results. With both results side-by-side it is easier to quickly compare the findings. This new feature can be found in the “InChorus Database Search”.
  • SPIDER (Software Protein Identifier) is now part of PEAKS. This sequence tag based search tool confidently identifies peptides using the de novo sequences derived by PEAKS auto de novo sequencing. In addition to exact sequence tag search, SPIDER can identify peptides from the database that show significant homology to the de novo sequence. Find protein information even in unstudied organisms, or find sequence variation between samples. SPIDER differs from regular homology search tools in that it was designed to account for inherent de novo sequencing errors like: (I/L), (N/GG), (SAT/TAS).
  • Thermo RAW data can be used directly without any conversion requirement. (XCalibur must be installed on the same computer).
  • The PEAKS algorithm has been modified to take better advantage of FT and LTQ OrbiTrap high mass accuracy data.
  • The Infochromics WIFF file data conversion tool can be plugged seamlessly into PEAKS Studio. Users with WIFF file data can now load and convert their data in one step.
  • To get the most out of lower resolution data, a set of data refinement tools have been created. Users can now filter out spectra of insufficient quality, recover peptide charge state information, and merge scans of the same peptide.
  • De novo and protein ID options panels have been reworked to provide better user accessibility. Enzyme and PTM information is readily accessible and editable from the main screen. The whole parameter set can now be saved and easily recalled for future use.
  • Taxonomy specific database queries with the Swiss-Prot database will speed up the search and improve the overall result quality by limiting the search to a smaller set of proteins.

  • Taxonomy specific database queries with NCBI NR will speed up the search and improve the overall result quality by limiting the search to a smaller set of proteins.
  • The Mac and Linux versions will be released with most of the 3.0/3.1 functionality — some 3rd party software that ships with PEAKS is unstable on MAC/Linux.

Performance enhancements

  • The PEAKS algorithm has been rewritten for version 3.1 with a 70% increase in speed in protein ID and auto de novo sequencing!
  • Consideration has been added for post-translational modifications (PTM) that occur only in the middle of a peptide.
  • The new mzXML file loader is 10 times faster.
  • De novo sequencing accuracy has been increased by 10%.
  • Charge state is calculated from the mzXML MS/MS data that yields better sequence interpretation.
  • The inChorus protein ID results give a more accurate confidence score by combining each protein search better.
  • The MS/MS spectrum merging tool has been improved to account for a series of successive scans of the same peptide over a large retention time range.

Interface enhancements

  • If retention time is present in the input data, it is stored as a spectrum property and is reported.

Compatibility with earlier versions

  • PEAKS 3.1 will read .ANN files generated by all earlier releases of PEAKS, and is fully forward-and-backward compatible with PEAKS versions 2.1 through 2.4. Files generated by PEAKS 3.1 can be read by PEAKS 2.0, but PEAKS 2.0 cannot make use of any embedded post-translational modification definitions. ANZ files generated by PEAKS 3.1 can be read by prior versions of PEAKS, but only 3.0 can make use of inChorus protein ID results.

Bugs fixes since last release

  • The XTandem Search part of the “inChorus” feature will not hang the analysis system if interrupted.
  • Various small bugs will be fixed that cropped up from the 3.0 version.

New features

  • inChorus protein ID technology allows users to run and automatically combine the results from several protein identification tools (PEAKS, TANDEM, OMSSA). This approach has been proven to significantly improve the result quality and coverage.
  • A high-throughput workflow creator makes processing several LC/MS/MS runs easy. Users can process a batch of data using whatever PEAKS tools they want.
  • For FTMS instruments ECD mode is now fully supported. The Spectrum Alignment View now displays c-ion and z-ion series alignment.
  • The Ion Table is now configurable to display any and all types of ions, ‘basic’ and ‘advanced’ defaults are provided.
  • Users are now able to seamlessly open Micromass/Waters .RAW data in PEAKS.
  • The ABI Converter, a data extraction tool for the ABI 4700 Oracle database has been developed, and now ships with PEAKS Studio. PEAKS Studio now reads the data files generated by the ABI Converter.
  • BSI better supports the mzXML file format and now reads in mzXML scan number and mzXML retention time information.
  • A ‘Merge Spectra’ tool has been added, allowing users to merge the MS/MS spectra of ions that represent the same peptide. This will reduce the amount of data to be processed, and improve its quality. Merging is dependent on similarity of precursor ion m/z, and retention time, if possible. Spectra may be merged only once, while loading, or after loading a data set.

Performance enhancements

  • With the inChorus technology, protein identification quality and coverage is much improved.
  • A new preprocessor has been developed for non-ion-trap data. As a result, users will see a 10% improvement in the results of auto de novo analysis on non-ion-trap data.
  • Improvements have been made to the PEAKS protein identification tool’s memory consumption.

Interface enhancements

  • A new reporting scheme has been implemented for use with inChorus protein ID. This shows the traditional protein report, plus a new peptide view – a sort-able list that tracks scan number, mass(calc), mass(exp), m/z, delta mass, sequence, protein accession, etc.
  • The Spectrum Alignment View now displays c-ion and z-ion series alignment.
  • he Ion Table is now configurable to display any and all types of ions, ‘basic’ and ‘advanced’ defaults are provided.
  • Users may now choose to preprocess their data, or not.

Compatibility with earlier versions

  • PEAKS 3.0 will read .ANN files generated by all earlier releases of PEAKS and is fully forward-and-backward compatible with PEAKS versions 2.1 through 2.4. ANZ files generated by PEAKS 3.0 can be read by PEAKS 2.0, but PEAKS 2.0 cannot make use of any embedded post-translational modification definitions. ANZ files generated by PEAKS 3.0 can be read by prior versions of PEAKS, but prior versions cannot make use of any inChorus protein ID results.

Bugs fixes since last release

  • Some N-terminus peptide modifications could not be found. Fixed.

New features

  • PEAKS has been fine tuned to take advantage of the superior accuracy and resolution provided by Fourier Transform (FTMS) instruments. CID, SORI, and IRMPD fragmentation modes are supported.
  • BSI is proud to support the proposed mzXML file format standard and thus Peaks will now open these files.
  • Peaks now records local time in the logging file to best assist technical support.
  • HTML reporting has been implemented. Export in HTML format allows import into certain spreadsheet programs for further analyses.
  • A mass calculator tool has been added, enabling the researcher to quickly calculate the mass of a peptide, including modifications.
  • A new report lists the spectra of quality not identified as part of a protein, and shows the de novo sequence found with each.

Performance enhancements

  • The accuracy and the confidence level evaluation has been improved on PEAKS auto de novo to better reflect the quality of results. The above two improvements have drastically increased the number of correct sequences, tags and amino acids obtainable by PEAKS auto de novo. Also as a result of the above improvements, PEAKS 2.4 is slightly slower at de novo sequencing.
  • Scalability of PEAKS Studio’s Protein ID capabilities has been improved. The user will no longer run out of memory on large datasets (our tests run up to 6000 spectra and have found no problems so far).
  • An error check has been added to the PTM editor. This lets the user know that the changes he/she is making to the current PTM.

Interface enhancements

  • A warning is shown when PEAKS fails to run because there is more than one copy running on a single machine.
  • The export peptide sequence to .FAS format function now exports the same score as the one shown in PEAKS.
  • The default spectrum and positional confidence colour scheme has been changed based on customer feedback.
  • Users may now use keyboard shortcuts to navigate through the Protein ID report, and can now copy text with CTRL+C.
  • The peptide sequences are now shown in Monospaced Sans Serif font: Lucidia Console.

Compatibility with earlier versions

  • PEAKS 2.3 will read .ANN files generated by all earlier releases of PEAKS, and is fully forward-and-backward compatible with PEAKS versions 2.1 through 2.3. ANN files generated by PEAKS 2.3 can be read by PEAKS 2.0, but PEAKS 2.0 cannot make use of any embedded post-translational modification definitions.

Bugs fixes since last release

  • In the PEAKS environment preference, color tab, the preview panel sometimes didn’t display because of differing JRE support. Fixed.
  • In the peptide data panel, drag and drop functionality (for example, creating a new node and drag-dropping some spectra there) failed. Fixed.
  • In the peptide data panel, right clicking selected a spectrum (and deselected all other spectra). Fixed.

New features

  • The formula for computing protein ID confidence has been changed to better distinguish between high-scoring proteins. Proteins require 5 times as much evidence as peptides, so e.g. 5 peptides at 90% confidence each will produce 90% protein confidence, 10 peptides at 90% each or 5 peptides at 99% each will produce 99% protein confidence, etc. A 5% confidence threshold is applied to remove some of the least likely proteins and keep the report a reasonable size.
  • PEAKS now correctly shows protein coverage information in the Protein ID result window. This coverage represents the number of amino acids found in sequence that match up to the sequence of the protein in question, and is expressed as a percentage of the length (in amino acids) of the whole protein sequence.
  • Protein identification results can be accessed by clicking on “Protein ID Result” beneath the filename in the Peptide Data Frame. This is in addition to the current navigation method.
  • The error tolerance input now allows free input, so you can experiment with high or low values for ion traps or FTMS (Note: for best results, use error tolerance values within the range 0.01 to 1.00 Da).
  • The database configuration wizard now provides FTP download links directly, or copies the link to the clipboard at the user’s choice. Performance enhancements

The following enhancements have in added into the PEAKS algorithm:

  • Speed and memory usage of Protein ID have been substantially improved.

Compatibility with earlier versions

  • PEAKS 2.3 will read .ANN files generated by all earlier releases of PEAKS, and is fully forward-and-backward compatible with PEAKS versions 2.1 through 2.3. ANN files generated by PEAKS 2.3 can be read by PEAKS 2.0, but PEAKS 2.0 cannot make use of any embedded post-translational modification definitions.

Bugs fixes since last release

  • In the Protein ID report, the protein coverage percentage has been corrected.
  • The erroneous positional confidences are now computed correctly.
  • Residues which are not present in the top scoring peptide may have non-zero positional confidence.
  • In the presence of many transpositions in the top sequences, positional confidence may be higher.
  • Protein ID uses the enzyme selected, and will no longer e.g. display non-tryptic results if tryptic digestion is selected.
  • Sparse spectra with no significant peaks no longer produce errors.

New features

  • Support for MicroMass ProteinLynx XML format. PEAKS can now open and process these files.
  • PEAKS now shows protein coverage information in the Protein ID result window. This coverage represents the number of amino acids found in sequence that match up to the sequence of the protein in question, and is expressed as a percentage of the length (in amino acids) of the whole protein sequence.
  • Protein identification results can be accessed by clicking a link in the main processing window of each spectrum. This is in addition to the current navigation method.
  • UniProt database format support and download URL are now provided. This, as well as the IPI databases can be used for species specific studies (Mouse, Rat and Human).

Performance enhancements

  • Major improvements to the quality of protein ID from Ion Trap spectra.
  • The number of proteins considered (internally) during database search has been increased – so as not to miss any protein candidates, and allowing more complex mixtures and larger numbers of redundant proteins in the database.
  • Selecting the enzyme Arg-C (or similar enzymes) now selects a fragmentation pattern which better accounts for the possible presence of Lysine.

Compatibility with earlier versions

  • PEAKS 2.2 will read .ANN files generated by all earlier releases of PEAKS. .ANN files generated by PEAKS 2.2 can be read by PEAKS 2.0, but PEAKS 2.0 cannot make use of any embedded post-translational modification definitions.

Bugs fixes since last release

  • Problems with manual sequencing commands have been resolved. Take special note of the repair to the search operations.

New features

  • The parameters that peaks uses to search a database and to perform auto de novo sequencing can now be defined separately. Also, database searching and auto de novo sequencing can be run independently of one another. The best practice is to perform de novo sequencing with only fixed PTM, then use variable PTM while searching the database.
  • When installing PEAKS, the user may now choose a 768M memory size option.
  • Any user-defined modifications found in sequence are saved in the .ANN file, for greater portability of results and if the user who doesn’t have that modification on their local machine, it can be imported it into their local configuration.
  • The “Database Configuration Wizard” assists in downloading configuring protein sequence databases so that protein ID result reports contain complete protein names and accession numbers. The user may choose to apply any database’s format to the results report. This can be selected during database configuration, or when viewing the report.
  • Peaks can now load datafiles in XML format, facilitating data transfer from ProteinLynx Global Server.

Performance enhancements

  • No major performance enhancements since last release.

Interface enhancements

  • Auto de novo and Protein Identification are given their own separate buttons.
  • The user can choose to run the database search separately from de novo or subsequently.
  • Independent parameters for each component are clearly presented / changeable.
  • Complete database configuration wizard assists in downloading configuring protein sequence databases so that protein ID result reports contain complete protein names and accesion numbers.

Compatibility with earlier versions

  • PEAKS 2.1 will read .ANN files generated by all earlier releases of PEAKS. .ANN files generated by PEAKS 2.1 can be read by PEAKS 2.0, but PEAKS 2.0 cannot make use of any embedded post-translational modification definitions.

Bugs fixes since last release

  • PEAKS no longer runs out of memory on certain datasets containing many unsequenceable spectra.

New features

  • PEAKS provides protein identification. Peaks can now use its de novo sequence results to aid in database searching and protein identification.
  • Wizards are provided to help users configure PEAKS.
  • PEAKS allows for differences in the quality of data between instruments. PEAKS now supports ion trap data.

Performance enhancements

  • Major accuracy improvement when sequencing using ion trap data.
  • Major accuracy improvement when sequencing non-tryptic peptides.

Compatibility with earlier versions

  • PEAKS 2.0 will read .ANN files generated by all earlier releases of PEAKS. PEAKS 2.0 .ANN files can be read by 1.x, but will lose any protein id information.

Bugs fixes since last release

  • Numerous minor bugs have been fixed.

  • Deep learning-boosted ID for PEAKS DB
  • Streamlined DIA workflow (Specral Library Search + directDB + de novo)
  • DeepNovo-based peptidome workflow for immunopeptidomics.
  • Quality Control (QC) Function for large cohort DDA and DIA
  • Support ZenoTOF
  • Other minor improvements and bug fixes

  • Redesigned backend using Cassandra database system
  • New web interface to submit jobs to the Online server
  • Performance scales linearly with hardware
  • Consistent results compared to PEAKS Studio
  • All search algorithms included in PEAKS Studio added to Online
  • Command line based interface available
  • Customisable workflows
  • Delimiter based sample adding
  • Batch exporting

Improvements to de novo Sequencing

  • Interactive Protein Coverage View
  • New Automatic de novo Result Validation
  • Improved de novo Residue Local Confidence Score
  • CID/HCD/ETD Complementary Fragmentation Support

New Label Free Quantification Module

  • Enhanced Feature Detection – Expectation Maximum (EM) algorithm
  • Significant improvement to -10LgP Score for Peptide Features and Proteins
  • New RT Alignment Algorithm
  • Normalization using TIC or selected protein(s)

Projects, Processing & Review

  • Improved speed for PEAKS DB, PEAKS PTM and SPIDER
  • Much improved data 2D/3D heatmap visualisation and navigation
  • New project wizard to facilitate the creation of complex project with a large number of samples and fractions
  • Reduced false discovery on PEAKS PTM algorithm

Collaborate

  • Better support for Thermo and Agilent raw files on 64bit OS
  • Pepxml exported from PEAKS can be imported into Skyline
  • Better support on mzidentML (version 1.1) format
  • Support pepxml exported from PD 1.4

Improvements to de novo Sequencing

  • Interactive Protein Coverage View
  • New Automatic de novo Result Validation
  • Improved de novo Residue Local Confidence Score
  • CID/HCD/ETD Complementary Fragmentation Suppor

New Label Free Quantification Module

  • Enhanced Feature Detection – Expectation Maximum (EM) algorithm
  • Significant improvement to -10LgP Score for Peptide Features and Proteins
  • New RT Alignment Algorithm
  • Normalization using TIC or selected protein(s)

Projects, Processing & Review

  • Improved speed for PEAKS DB, PEAKS PTM and SPIDER
  • Much improved data 2D/3D heatmap visualization and navigation
  • New project wizard to facilitate the creation of complex project with a large number of samples and fractions
  • Reduced false discovery on PEAKS PTM algorithm

Collaborate

  • Better support for Thermo and Agilent raw files on 64bit OS
  • Pepxml exported from PEAKS can be imported into Skyline
  • Better support on mzidentML (version 1.1) format
  • Support pepxml exported from PD 1.4
  • Improved 3rd party exporting to Scaffold and resolved PTM issue
  • Fixed mzIdentML protein reporting issue
  • Fixed a regression bug affecting charge and mass correction on high charge Orbitrap data
  • Third party and pepXML export from inChorus results are now disabled. It was unintentionally opened in the initial PEAKS 7 release. In mzIdentML, there are currently no CV term in the standard for PEAKS inChorus results
  • Unique peptide filter in Label Free Quantification results now correctly uses the unique peptides instead of the peptide features as the filter
  • Fixed an issue in Label Free Quantification p-value calculation. Users do not need to redo the analysis. Re-open the results will have the p-value displayed correctly
  • PEAKS now supports 64 bit CompassXtract libraries from Bruker
  • PEAKS now can read the scan number from MGF files generated by the new Compass tools
  • Better handling on MGF/PKL/DTA files with excessive white spaces
  • Better handling on importing Mascot results into PEAKS. Solved issues for some custom FASTA header formats
  • Resolved slowe Derby initialization issue
  • PEAKS project converter: MS1 spectrum information is correctly converted from PEAKS 6 projects
  • Added Mascot 2.5 support in inChorus
  • Better support on Waters raw data
  • Fixed an issue related to usage when using a different user account
  • Fixed an issue in mzIdentML exporting when some fractions had no results
  • Fixed an issue reading Bruker raw data
  • Fixed an issue that caused OMSSA search to sometimes fail after a Mascot search

Algorithm

  • New nonspecific enzyme digestion support. Now one can allow nonspecific enzyme digestion at 0, 1, or both ends of a peptide. Having more nonspecific digestion ends will increase the search sensitivity at a reduced search speed
  • Consider neutral loss of phosphorylation and sulfation PTMs in the scoring functions of PEAKS DB and PEAKS PTMs
  • New -10lgP score and FDR control for PTM and SPIDER results
  • FDR control of the inChorus result. Different search engines results can now be combined according to a unified FDR for each search engine
  • Improved precursor mass correction. The mass correction function in “data refine” step recognizes the real monoisotopic mass even if the instrument’s raw data reported the isotope mass. The algorithm is improved in the new version

GUI Improvements

  • Easier PTM selection interface. The PTMs are separated in three “common”, “uncommon”, “customized” to make it easy to specify the PTMs to search for. Additionally, a “Recent” list includes all the PTMs recently used by the user
  • De novo only view added to inChorus result. The de novo only view reports peptides found exclusively by de novo sequencing. This useful view is now available to the inChorus result too
  • Improved spectrum annotation view. Mouse over an amino acid in the spectrum annotation will highlight the supporting fragment ions. Setting an anchor peak will show mass different between the current peak and the anchor peak
  • Improved project creation interface. The selection of instrument type and fragmentation mode is easier. Users can specify a proteolysis enzyme for each sample at the project creation interface
  • More searching functions to locate a specific de novo sequence result. Now de novo sequencing results can be sorted and searched with the amino acid sequence
  • Display absolute intensity in PEAKS Q. Users now can choose to display absolute intensities or ratios in PEAKS Q results. For ratio display, users can choose which sample the ratio is based
  • More statistics in the summary view. The PTM profile table is improved. A protein FDR value is added
  • One-click specification of common FDR values (e.g. 1%) in the FDR selection pane

Better Community Support

  • A new configuration wizard to assist the download and installation of public protein databases and raw file readers/converters
  • Major accuracy and sensitivity improvement on the analysis of AB SCIEX TripleTOF data
  • Proteome Discover support. Now PEAKS can load the pepXML result file generated from Thermo’s Proteome Discover software
  • Support mzIdentML result format. Now PEAKS can export mzIdentML file for downstream analysis such as Scaffold PTM
  • Export high resolution images of the spectrum annotation and protein coverage view
  • Export to website or single webpage format for easy sharing of results

Unprecedented Accuracy and Sensitivity for Database Search

  • The revamped database search engine (PEAKS DB) results in substantially improved accuracy and sensitivity for peptide identification, causing the identification of more peptides with a reduced false discovery rate (FDR). In particular, the preview version of the PEAKS DB engine produced excellent performance in the ABRF/iPRG 2011 study for ETD data analysis

Comprehensive Result Visualization

  • Numerous improvements were made to support the visual examination of the results from all different angles. To enhance user confidence, PEAKS provides tools for vigorous yet convenient result examination
  • Users will also notice many areas where a summary of the result’s statistics are automatically provided. Through these charts users can easily answer important questions such as whether the target-decoy FDR calculation is reliable, and whether the instrument is well-calibrated. Additionally, the peptide and protein tables were rewritten to provide convenient search and sorting functions, allowing easy location and examination of certain particularly interesting peptides or proteins

Built-in Result Validation

  • PEAKS DB provides a seamlessly built-in result validation system with an enhanced target decoy method. Score thresholds can be conveniently selected from the FDR curve, avoiding the guess work performed with other software

Enhanced result reporting

  • The new summary view provides a central place for specifying score thresholds to filter results. Results can be filtered quickly and efficiently with the summary view showing the changes at a glance. Filtered results can easily be exported to several CSV and HTML files for publication and/or result sharing with non-PEAKS users. Furthermore, results can be saved as a PEAKS project to be viewed by your collaborators with a free PEAKS Viewer

De novo sequence tag generation

  • By specifying a confidence threshold with a user-friendly sliding bar, users can promptly convert the acclaimed PEAKS de novo sequencing results into high confidence de novo tags. Users can even export these tags for integration with their own in-house analysis workflow

Improved inChorus search

  • The inChorus function combines the search results of several search engines, including Mascot 2.3. An intuitive search engine icon displays which specific engines identified each peptide and when multiple engines agree. The new filtration rule allows users to filter the results flexibly, including by individual engine scores. A Venn diagram and side-by-side FDR curves display each engine’s contribution to the final result

Heatmap view of quantification results

  • The heatmap provides a bird’s-eye view of the protein quantity changes across different groups of samples. The proteins are automatically clustered according to their quantity change patterns, facilitating the quick identification of the possibly interesting proteins and patterns

Heatmap view for LC-MS data

  • The data heatmap provides a bird’s-eye view of the peaks in the LC-MS data, with peptide features highlighted

New statistical peptide score

  • The scoring function for evaluating peptide-spectrum matches is significantly improved in 5.3. This is the foundation of the increased accuracy and sensitivity of PEAKS DB. The scoring schema of peptide identification involves matched peaks and their intensities, precursor mass error, enzyme specificity, de novo sequence, peptide size, and more. A statistical evaluation, -10lgP, is given for each peptide-spectrum match. Here lg() is the common logarithm with base 10, and P is the probability that a false identification of the current search has the same or better significance

Algorithm

  • Overall improvement of software capacity, stability and speed
  • Increased de novo and database search accuracy for ETD data
  • Automatic precursor mass and charge correction for high resolution data
  • Improved the SPIDER scoring function
  • Reduced the memory overhead for SPIDER, improving its capacity and speed
  • PTM support in manual de novo
  • Improved workflows targeting different application needs such as Identification, Quantification and inChorus
  • Replicate analysis for label free quantification
  • FASTA database validation, important for custom databases
  • CD/HCD, CID/PQD support in iTRAQ quantification with Thermo instruments
  • Improved support for Thermo LTQ RAW file format
  • mzML/mzXML3.0/mzData/pepXML file format support

Analysis

  • MCP compliant result exporting
  • Result comparison interface
  • Improved spectrum display
  • Standardized PTM display
  • Optional display of assigned/unassigned de novo peptides in PEAKS Search results
  • inChorus Search parameter saving

  • Industry grade security and user management: System administrators can empower users and research groups to share resources without compromising security
  • Efficient collaboration: Easily and automatically send/share results with clients/customers and coworkers
  • 10% – 30% improvement in de novo sequencing accuracy on ion-trap data
  • 11 instrument selection for maximum flexibility
  • Improved charge and monoisotopic peak picking mean even better coverage
  • New GUI design for optimal function and navigation

Maximized User Privileges:

  • Individual user account, complete with password protection
  • Self-define individual databases, enzymes, post translational modifications
  • PPM support for parent mass error tolerance

Maximized Administrative Capabilities:

  • Ultimate control in defining user privileges
  • Configuration modifications are made easily in the GUI, instantaneously updated across the network

New Features

  • Taxonomy specific database queries with Swiss Prot will speed up the search times by narrowing the number of records that are compared and also narrowing the results to specific taxonomies
  • The node task allocation has been optimized to increase individual cluster node performance. The optimization ensures that all the nodes are as busy as possible which increases overall cluster performance
  • Data refinement has been incorporated with PEAKS Online. Charge recognition, low quality spectra filter and preprocessing functions have been added to get the most out of your data
  • The result display now include the scan number and retention time if the data is present in the mzXML input file
  • Any inputted zipped DTA file now have their names displayed in the results

New Features

  • Taxonomy specific database queries with NCBI NR will speed up the search times by narrowing the number of records that are compared and also narrowing the results to specific taxonomies
  • User can define custom PTM’s to be used in an analysis
  • The new “Peptide view” in the resulting .anz file can be used to sort the peptide results quickly and efficiently

Performance Enhancements

  • The protein ID search algorithm has improved scalability. A single task can contain tens of thousands of spectra
  • The speed of the auto de novo and protein identification has been doubled. This means that the overall performance of Online is approximately 4 times faster

Compatibility with earlier versions

  • The results and .anz files produced by PEAKS Online 1.0 are all compatible to the new results format found in PEAKS Online 1.1

Bugs fixes since last release

  • The “outofboundary” exception bug that occurs when the client defined database is very small has been fixed

Features

  • PEAKS Online is a cluster server with a web interface that allows multiple users to share data and processing power. This product has all the power of PEAKS Studio with the added benefit of a cluster server. A cluster server allows multiple PCs to work together on the same data set. This cuts processing time to a fraction of what it would take with one system doing all the work
  • PEAKS Online has high availability. PEAKS Online is a web server for Peaks which means users can always use PEAKS Online, no matter where she/he is, as long as the user has internet connection and web browser
  • Sharing your data results with trusted colleagues and clients is easy with the help of a browser and an internet connection
  • The web interface is a single page that focuses on browsing and analysis options. Being so simple, users of the software can begin to analyze data and obtain meaningful results in no time
  • PEAKS Online will send the results by email. The results are in both .html and .anz formats. The results were saved on the server so the user can always access the results as long as they are not deleted by the software administrator
  • The result in the .html format can be sorted by m/z , sequence or score. (Note. only for data with less than 500 spectra because for big data it will take a very long time to load the page)
  • PEAKS Online has a task queue maintained on the database server. Users can always submit the tasks which will be put into the end of task queue automatically if the PEAKS Online server is busy with other task. Even the PEAKS Online server is down for some reason the task queue will not be destroyed or lose any task(s)
  • PEAKS Online can be configured to use one of several task schedule schemes. For example, heavy duty tasks can be scheduled to be processed during a certain time of the day only, which will make the server more responsive to normal sized tasks
  • PEAKS Online offers a free online Mass Calculator for all PEAKS Online’s customers