Complexomics: MitCOM

This package contain supplementary code for the MitCOM publication.

Content

Package

The mitcom package contains all relevant methods for handling protein profiles and single profile components: from component detection and correlation to similarity distance matrix calculation.

Peak detection script

The complex_find_peaks executable for detecting peaks (aka components) in protein profile data is available after package installation.

It expects at least two tab-separated (tsv) input files with the following structure:

File 1: Protein abundance values

Protein	Sample-1	Sample-N-1	Sample-N
Prot-1	...	...	...
Prot-2	...	...	...
Prot-3	...	...	...

File 2: Apparent molecular weight values

The file may contain any number of columns, only the first and the MW column will be read. Values in the first column must match headers in File 1 above.

Sample	...	MW
Sample-1	...	350.3
Sample-2	...	423.6
...	...	...
Sample-N	...	3244.5

Parameters:

You can trigger plot creation by specifying a plot output directory, set the number of parallel workers, specify the output file names, a protein whitelist and more. Run complex_find_peaks --help to see all options.

Notebook

The MitCOM.ipynb notebook exemplifies the way from data import to t-SNE output. It requires the mitcom package to be installed.

Data files loaded from notebook

• protein_abundance_file: Protein abundances (see above)

• components_file: Component output file from peak detection script

• protein_whitelist_file: Text file containing whitelisted proteins (optional)

• predefined_complexes_file: Two-column text file (tsv). Can be used to define known complexes by mapping component identifiers to arbitrary complex names. This mapping is only used during t-SNE visualization. Data points belonging to the same complex will be plotted as one individual trace that can be shown/hidden by clicking on its associated legend item. Note: The file is expected to exist and contain at least the header row.

  Example:

  component    complex
  ATPA_P6      ATP#1
  ATPB_P6      ATP#1
  ATPD_P3      ATP#1
  SDH3_P8      SDH
  SDHA_P8      SDH
  SDHB_P9      SDH
  ...          ...
  

Installation

Standard

# optional: create and activate virtual environment
python -m venv .venv && .venv/Scripts/activate

# optional: install wheels to speed up installations
pip install wheels

pip install git+https://github.com/phys2/complexomics-mitcom.git@main

With Jupyter notebook support

The optional dependency group notebook contains additional packages for running the Jupyter notebook, namely jupyter, jupyterlab and plotly. To have these included, run these commands instead:

pip install "git+https://github.com/phys2/complexomics-mitcom.git@main#egg=complexomics-mitcom[notebook]"

# download MitCOM notebook
curl -L -O https://github.com/phys2/complexomics-mitcom/raw/main/mitcom/notebooks/MitCOM.ipynb