https://github.com/vlopezferrando/pymut

Machine learning methods for the prediction of pathology in protein mutations, as in the PMut predictor.
https://github.com/vlopezferrando/pymut

machine-learning mutations pathology protein-mutations

Last synced: about 1 year ago
JSON representation

Machine learning methods for the prediction of pathology in protein mutations, as in the PMut predictor.

• Host: GitHub
• URL: https://github.com/vlopezferrando/pymut
• Owner: vlopezferrando
• License: mit
• Created: 2017-02-10T10:03:13.000Z (over 9 years ago)
• Default Branch: master
• Last Pushed: 2018-01-08T10:27:43.000Z (over 8 years ago)
• Last Synced: 2025-03-24T02:22:07.370Z (over 1 year ago)
• Topics: machine-learning, mutations, pathology, protein-mutations
• Language: Python
• Homepage: https://mmb.irbbarcelona.org/PMut/pymut
• Size: 50.6 MB
• Stars: 6
• Watchers: 2
• Forks: 4
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE.txt

Awesome Lists containing this project

README

          

# PyMut

PyMut is a Python 3 module that fills the gap between machine learning and

bioinformatics, providing methods that help in the prediction of pathology in

protein mutations. Using PyMut you can compute features, train predictors,

evaluate them, predict the pathology of mutations, select the best features...

## Installation

We recommend using Anaconda, a free distribution of the SciPy stack. To install

PyMut, follow these steps:

1. 

Download anaconda with Python 3.5

for your platform (Windows, Linux or OSX), and install (double-click in Windows), or run in Linux of OSX:

    `bash Anaconda3-4.3.0-Linux-x86_64.sh`

2. Create an anaconda Python 3 environment (we will name it pymut), and activate the

environment:

    `conda create python=3 --name pymut`

    `source activate pymut`

3. Install PyMut:

    `pip install pymut`

## Tutorial

Visit the



PyMut tutorial

to see a full example of usage of PyMut. In the tutorial we show how to:

* Compute features that describe mutations and plot their distribution.

* Train classifiers, evaluate them using cross-validation and plot their ROC curves.

* Select the best features.

* Train a pathology predictor.

* Predict the pathology of mutations using our newly trained predictor.

## Sample plots generated by PyMut