Online Tools/Calculators

• MS/MS fragmentation calculator
• Protein sequence digestion calculator
• Isotope distribution calculator
• Elemental mass calculator
• Lorikeet Spectrum viewer
• Table of relevant masses (amino acids, elements)

MS/MS database search tools

• Comet (local linux commands)
• X!Tandem
• Crux (local run info)

Software tools and protocols from other resources:

• Skyline: targeted proteomics environment
• XLink-DB: database and software tools for storing and visualizing protein interaction topology data
• Proxl Protein Cross-Linking Database

CrossLinking

• ProXL
• XlinkDB

Miscellaneous

• A quick primer on how to interpret peptide ID results processed at UWPR
• Simple .ms2 to .dta command line converter (zip of source + binaries)
• Simple .ms2 to .mgf command line converter (C source)
• Here are TPP Libra condition files for iTRAQ and TMT.
• ReAdW under Wine Thermo to mzXML conversions under linux.
• RawMeat is a nice Thermo raw file diagnostic tool developed by the now defunct Vast Scientific. RawMeat requires Thermo's MSFileReader library to be installed. MSFileReader can be downloaded by creating an account with ThermoFischer Scientific. Once registered you will find the software under "Other Software Releases".
• Instructions on running Abacus spectral count tool on UWPR systems.
• .tgz (tar-gzip) archive of contaminant protein sequences based on the GPM's cRAP database. Updated 2022/06/21 to current UniProt sequences.

MS/MS database search info:

• Search engine comparisons: link1, link2, link3 (only OMSSA searched 3+ spectra!), link4, link5, link6
• Be wary of searching with too narrow a precursor tolerance (even if your mass spectrometer can measure that accurately). It's an issue with all search tools. This paper nicely touches on the problem and here's a rebuttal. And Phil Wilmarth has a nice blog post about the benefits of using a wide precursor window.
• Be wary of multiple-stage searches like X!Tandem's refinement mode or the hot-spot searching that Paragon does. Not that these searches are innately bad but rather one should be educated on the search strategy and what it means. If one searches a bazillion spectra against subset proteins X, Y, and Z in the 2nd pass, don't be surprised to find many matches (real or spurrious) to proteins X, Y and Z. Think of it as a self-fulfulling prophecy. Here's an interesting paper and follow-up touching on the problem and proposing solutions to the target-decoy strategy with these types of searches: paper1, paper2

UWPR computing resources (as of 10/2024)

• Two (2) Dell PowerEdge R6515 servers each with 256GB RAM, 480GB SSD, and dual AMD EPYC 7443P 2.85GHz CPUs (each with 24 cores, 48 threads, 128M Cache). These serve as cluster compute nodes (p001 and p002).
• One (1) Dell PowerEdge R660 server with 128GB RAM, Intel Xeon Silver 4510 2.4GHz CPU (12 core, 24 threads, 30M Cache). This serves as an application and web server (tephra).
• 23TB of enterprise class storage with NFS (Linux) and SMB (Windows) connectivity.
• The UWPR also has access to UW Genome Sciences cluster computing resources. This includes (A) one compute node with dual Intel Xeon Gold 6330N 2.2GHz CPUs (each with 28 cores, 56 threads, 42MB cache), 512GB RAM, 890GB local storage; (B) two compute nodes with dual Intel Xeon Silver 4316 2.3GHz CPUs (each with 20 cores, 40 threads, 30MB cache), 256GB RAM, 890GB local storage; and (C) three compute nodes with Intel Xeon E5-2680 2.5GHz CPUs (8 cores, 16 threads, 20MB cache), 256GB RAM, 930GB local storage.