Data

Tools

Indexes

Applications

Experiments

Open Source Science

ksfinder

This repository holds code & data for KSFinder, a kinase-substrate link prediction tool

https://github.com/manju-anandakrishnan/ksfinder

Science Score: 57.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ✓
    DOI references
    Found 2 DOI reference(s) in README
  • ○
    Academic publication links
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (9.4%) to scientific vocabulary
Last synced: 6 months ago · JSON representation ·

Repository

This repository holds code & data for KSFinder, a kinase-substrate link prediction tool

Basic Info
  • Host: GitHub
  • Owner: manju-anandakrishnan
  • License: other
  • Language: Python
  • Default Branch: main
  • Homepage:
  • Size: 5.26 MB
Statistics
  • Stars: 0
  • Watchers: 0
  • Forks: 0
  • Open Issues: 0
  • Releases: 4
Created almost 3 years ago · Last pushed about 1 year ago
Metadata Files
Readme License Citation

README.md

git clone https://github.com/manju-anandakrishnan/ksfinder.git
cd ksfinder

Create environment and load libraries

conda create --name ksfenv python=3.7
conda activate ksfenv
conda install --file requirements.txt
pip install ampligraph
pip install ordered_set
pip install suds
conda install openpyxl

Hardware Requirements - Nvidia GPU with at least 16 GB memory and support for CUDA 10.0 or higher.

Initialize & download the data from Zenodo

bash init.sh
export PYTHONPATH=$(pwd)

Load KG data & assess the model (Assessment 1 & 2)

python preprocess/src/main.py
python kge/src/main.py

To train knowledge graph embedding models from scratch, pass the argument --t_kge=True to the above script (KGE embedding from scratch may take weeks to complete depending on the GPU capability. Optionally use the trained KGE models)

The results of evaluation and test data will be loaded in the assessment folders 1 & 2.

Train the MLP classifier and assess KSFinder (Assessment 3 & 4)

python clf/src/main.py

To train the classifier model with the embedded vectors, pass the argument --t_clf=True to the above script. The results of evaluation and test data will be loaded in the assessment folders 3 & 4.

Comparative assessments (Assessment 5, 6, 7, 8, 9)

python compare/src/main.py
The results of evaluation and test data will be loaded in the appropriate assessment folders.

Predict using KSFinder

python ksfinder/src/main.py
The prediction results will be loaded in the folder, ksfinder/results

Literature mining using iTextMine API

python textmine/src/main.py

Enrichment Analysis using the predicted substrates

python enrich_analysis/src/main.py

By using the data or code in this repository, you accept the terms and conditions governed by the license.

If you use any part of our work/data, please cite us:

Anandakrishnan M, Ross KE, Chen C, Shanker V, Cowart J, Wu CH. 2023. KSFinder—a knowledge graph model for link prediction of novel phosphorylated substrates of kinases. PeerJ 11:e16164 https://doi.org/10.7717/peerj.16164

Owner

  • Name: ManjuA
  • Login: manju-anandakrishnan
  • Kind: user

Citation (CITATION.cff) 

If you use this data or code, cite the paper,
Anandakrishnan M, Ross KE, Chen C, Shanker V, Cowart J, Wu CH. 2023. KSFinder—a knowledge graph model for link prediction of novel phosphorylated substrates of kinases. PeerJ 11:e16164 https://doi.org/10.7717/peerj.16164

GitHub Events

Total
  • Push event: 1
Last Year
  • Push event: 1

Dependencies

requirements.txt pypi
  • cudatoolkit *
  • cudnn *
  • matplotlib ==3.5.3
  • numpy ==1.21.5
  • pandas ==1.3.5
  • requests ==2.28.1
  • scikit-image ==0.19.3
  • scikit-learn ==1.0.2
  • tensorflow-gpu ==1.15