cobia is a computational tool to predict the number of cofragmenting ions in a mass spectrometry experiment, focused on metaproteomics. Specifically, we calculate 'cofragmentation scores' which represent identification and quantification bias using mass spectrometry based metaproteomics.
I've removed the ability to use BioLCCC due to consistent and unclear errors. I've also updated the targeted approach, so that if you have a peptide in your targeted file that's not in your database, it doesn't stop the prediction, it just doesn't provide metrics of cofragmentation.
The installation instructions assume a linux environment, although they can be adapted to work on a Windows or MacOS system as needed. The following python modules are required:
pandas
version > 0.2.0pyteomics
- Retention time prediction:
BioLCCC
or RTModel/RTPredict from OpenMS
- Python 2.7 (for compatability with BiolCCC)
pyopenms
(for subsampling retention times from an idXML file)
To install cobia, download the source code, and in this directory run the
setup.py
file. Running the following command will build and install cobia:
python setup.py install
I recommend installing on a conda
environment:
python setup.py install --user
Here is a conda environment command that would setup your environment:
conda create --name pyteo_27 python=2.7 pandas
conda activate pyteo_27
pip install pyteomics.biolccc
pip install pyteomics
pip install pyopenms
A single line fasta file, predicted retention times (recommend using RTModel or another retention time predictor, which requires an mzML file from an MS experiment).
A csv file of peptides and cofragmentation scores.
To use cobia, please see the how-to above! (in the 'Writing' folder)
Convert over to python 3.6: Originally, I had used python 2 because I wanted to use BioLCCC. I've been getting some strange errors using BioLCCC with memory. Also, I think the 2nd approach in the manusript is more flexible (training a retention time model off of observed peptides).