- Free software: MIT license
- Official Documentation available at: https://msci.readthedocs.io.
Peptide identification by mass spectrometry relies on the interpretation of fragmentation spectra based on the m/z pattern, relative intensities, and retention time (RT). Given a proteome, we wondered how many peptides generate very similar fragmentation spectra with current MS methods. MSCI is a Python package built to assess the information content of peptide fragmentation spectra, we aimed calculating an information-content index for all peptides in a given proteome would enable us to design data acquisition and data analysis strategies that generate and prioritize the most informative fragment ions to be queried for peptide quantification.
prerequisites:
- Python 3.8 -3.11
- Anaconda
- Matchms
Here is a small example of using MSCI to calculate pairwise normalized spectral angle .. testcode:
from MSCI.Preprocessing_data import read_msp_file from MSCI.grouping.groups import MassContentInformation, process_data from MSCI.similarity.Similarity import joinPeaks, nspectraangle from MSCI.utils import process_combin, parallel_function import numpy as np import multiprocessing as mp from functools import partial from multiprocessing import Pool, cpu_count File= 'MSCA_Package/Tryptic_peptides/Dataset/msp_files/charge2_3myPrositLib.msp' mz_irt_df = read_msp_file(File) g = MassContentInformation(mz_irt_df) group = g.group_sequences(1,10, unit='Da') group = np.array(group, dtype=object) combin = process_data(group) # Create a partial function of process_combin with relevant_spectra and other parameters process_combin_partial = partial(process_combin, spectra=relevant_spectra, tolerance=1, ppm=0) # Determine the number of CPU cores available num_cores = cpu_count() # Use a Pool to parallelize the processing with Pool(num_cores) as pool: results = pool.map(parallel_function, updated_combin_chunk)
Should output a list of peptides and their spectral angles
You can install MSCI via pip_ from PyPI_:
$ pip install MSCI
If you would like to contribute to this project, feel free to fork the repository on GitHub and submit a pull request.
This package was created with cookietemple_