Skip to content

Repository files navigation

GeoStab

Improving the prediction of protein stability changes upon mutations by geometric learning and a pre-training strategy

Data

Dataset/Database Type Composition url Time
DeepSequence dataset Fitness 34 single-point mutation DMS studies 2018-09
MaveDB Fitness The database contains fitness data of different mutations https://www.mavedb.org/ 2022-09
GeoFitness dataset Fitness 74 DMS datasets dms.csv 2023-05
ProThermDB ΔΔG and ΔTm The database contains ΔΔG and ΔTm data of different mutations https://web.iitm.ac.in/bioinfo2/prothermdb/index.html 2021-07
ThermoMutDB ΔΔG and ΔTm The database contains ΔΔG and ΔTm data of different mutations https://biosig.lab.uq.edu.au/thermomutdb/ 2021-07
S2648 ΔΔG 2648 single-point mutation from 131 proteins This file is included in S8754.csv 2009-08
S8754 ΔΔG 8754 single-point mutation from 301 proteins S8754.csv 2023-05
S783 ΔΔG 783 single-point mutation from 55 proteins S783.csv 2023-05
S669 ΔΔG 669 single-point mutation from 94 proteins S669.csv 2022-01
S461 ΔΔG 461 single-point mutation from 48 proteins S461.csv 2023-01
S1626 ΔTm 1626 single-point mutation from 95 proteins This file is included in S4346.csv 2016-06
S4346 ΔTm 4346 single-point mutation from 349 proteins S4346.csv 2023-05
S571 ΔTm 571 single-point mutation from 37 proteins S571.csv 2023-05

Other data files in the data folder

Prerequisites

Install software on Linux

  1. download GeoStab
git clone https://github.com/Gonglab-THU/GeoStab.git
cd GeoStab
  1. install FoldX software

  2. install Anaconda software

  3. install Python packages from Anaconda

conda create -n geostab python=3.10
conda activate geostab

conda install pytorch cpuonly -c pytorch
pip install biopandas
pip install biopython
pip install click
pip install pdb-tools

Usage

  • You should modify the mutant information in generate_example_mut_info_csv.py (0 <= pH <= 11 and 0 <= temperature <= 120)
  • You should modify the software_foldx in run.sh
  • You should install AlphaFold2 and modify script in run.sh to generate Seq version structure
# fitness Seq prediction
python generate_example_mut_info_csv.py
bash run.sh -m fitness -v Seq -o ./example_Seq

# fitness 3D prediction
python generate_example_mut_info_csv.py
bash run.sh -m fitness -v threeD -o ./example_3D

# ΔΔG/ΔTm Seq prediction
python generate_example_mut_info_csv.py
bash run.sh -m ddGdTm -v Seq -o ./example_Seq

# ΔΔG/ΔTm 3D prediction
python generate_example_mut_info_csv.py
bash run.sh -m ddGdTm -v threeD -o ./example_3D

Reference

Improving the prediction of protein stability changes upon mutations by geometric learning and a pre-training strategy

Acknowledgements

About

No description, website, or topics provided.

Resources

Stars

39 stars

Watchers

3 watching

Forks

Releases

Packages

Contributors

Languages