pasrawin / libraryproteomicmsd

The extension of https://github.com/pasrawin/ProteomicNMF/

Non-negative matrix decomposition for proteomic mass spectrogram using a library-based dictionary

The proteome-scale application of https://github.com/pasrawin/ProteomicMSD/

Overview

Library searching has been extensively developed to improve the identification speed, accuracy, and sensitivity. We applied proteomic MSD with peptide detectability prior knowledge as the concept of a library-based dictionary for proteome-scale data.

Requirements

Proteomic Mass Spectrogram Decomposition (protMSD) was written in Python 2.7 and tested on Window systems.

Dependencies

• division from future, numpy, pandas, scipy, math, collections, itertools, sklearn, gc, and pyteomics

The inputs are:

1. An .mzML or a .txt file of LC/MS experiment converted by ProteoWizard MSconvert, or a .pkl file of protMSD output (required)
   • Example: Example_protNMF_Ecoli_msconvert.txt (available here)
     • After running protMSD, Example_protNMF_Ecoli_msconvert.pkl will be obtained. You may reuse this in following protMSD run to save the computational time.
2. An .xlsx of identification results with retention times for generating a library (required)
   • Example: Example_protNMF_Ecoli_mascot.xlsx (available here)
     • After running protMSD, Example_protNMF_Ecoli_insilico.xlsx will be obtained. You may reuse this in following protMSD run to save the computational time.
   • Example: Example_protNMF_Ecoli_insilico.xlsx (available here)

The outputs are:

1. A .pkl file of LC/MS experiment
2. An .xlsx file of MSD result
   • Worksheets: 1.XIC, 2.Peakresult
3. Three .npz files of V, W, and H matrices

Installation

git clone https://github.com/pasrawin/LibraryProteomicMSD.git

Running

1. Prepare a directory containing your input files
2. Define your input file names and parameters in protNMF00_handler.py
   • The mass spectrograms from different instruments and proteomic experiments provide different features. In order to obtain an optimal result from protMSD, we strongly suggest that the parameters should be carefully set for each observed mass spectrogram.
   • The m/z range, retention time range, smoothing window and shift, bins, in silico digestion criteria, number of best peaks reported etc. can be modified easily by replacing default values here.
3. Run python mNMF00_handler.py
   • Yor command prompt will show protMSD process from 1 to 10

Benchmark Datasets

The MS raw data were deposited at the ProteomeXchange Consortium via jPOST partner repository with identifier JPST000765. Currently, available for reviewers only. Please use the access key in Supplementary Information).

Support

If you have any questions about protMSD, please contact Pasrawin Taechawattananant ([email protected]), Kazuyoshi Yoshii ([email protected]), or Yasushi Ishihama ([email protected])