Skip to content

PTM-X: a prediction tool for Post-Translational Modification cross-talk

License

Notifications You must be signed in to change notification settings

huangyh09/PTM-X

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

38 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PTM-X: PTM cross-talk predictor

About PTM-X

PTM-X is an algorithm to predict post-translational modification(PTM) cross-talk both intra- and inter- protein. In current version, the following four features are used to predict PTM cross-talk inter proteins, via a random forest classifier.

1. sequence residue co-evolution: The coevolution between two the amino acids across multiple vertebrates. The multiple sequence alignment data is downloaded from eggNOG v4.5. Here, the 1-Hamming loss is used to measure the co-evolution score.

2. sequence motif co-evolution: The +/-3 surrounding amino acids are used to form the motif for a given PTM locus. The fraction of consistence the this motif in any vertebrate comparing to its human orthologous protein. Based on the two vectors of motif conservation fraction, the co-conservation score is calculated by the mean product of the two vectors. Again, the eggNOG v4.5 is used here.

3. Co-modification across different species: The co-conservation of PTM existence across human, mouse and rat. The raw PTM data for these three species is downloaded from PhosphoSitePlus.

4. Co-modification across different conditions: The co-occurrence between the two PTMs across 88 tissue, disease and cellline conditions.

Get started

Installation

Step 1: you need a Python environment for supporting packages. The easiest way might be installing the python platform via Anaconda. PTM-X is only compatible with Python 2.7. If you are using Anaconda 3, create a conda environment with Python 2.7 and dependent packages as follows,

conda create --name ptmxPy2 python=2.7 numpy scipy scikit-learn==0.17 joblib==0.11
# activate the environment
conda activate ptmxPy2

Step 2: install PTM-X package by either downloading the source code here or use the following command line for an equivalent setting:

pip install -U git+https://github.com/huangyh09/PTM-X

In either case, if you don't have the root permission, add --user.

Step 3, to fetch above four featuresd ownload the data from http://ufpr.dl.sourceforge.net/project/ptm-crosstalk/PTM-X_data_v2.2.zip Unzip it and set the directory as the according parameter in the command lines below.

Fetch features

  • To see all arguments, try command line PTMX-feature -h
  • For the data_dir, set it as the downloaded data (see above).
  • The python file is feature_extractor.py

Prediction

  • To see all arguments, try command line PTMX-predict -h
  • The python file is predict.py.

Links

References

About

PTM-X: a prediction tool for Post-Translational Modification cross-talk

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published