PremPS

PremPS evaluates the effects of single mutations on protein stability by calculating the changes in unfolding Gibbs free energy. It can be applied to a large number of tasks, including finding functionally important variants, understanding their molecular mechanisms and protein design. 3D structure of a protein is required for this method.

We recommend that most users who just want to obtain PremPS predictions use PremPS website to obtain scores.

You can download releases on github.

PremPS requires the following software and packages.

1. DSSP

   This is available at the DSSP website.

   https://swift.cmbi.umcn.nl/gv/dssp/

2. PROVEAN

   This is available at the PROVEAN website.

   http://provean.jcvi.org/index.php/

3. NCBI BLAST 2.4.0

   This is available at the NCBI ftp site.

   ftp://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/2.4.0/

4. FoldX

   This is available at the FoldX website.

   http://foldxsuite.crg.eu/

5. VMD

   This is available at the VMD website.

   https://www.ks.uiuc.edu/Research/vmd/

6. Python packages: pandas and rpy2

   To install these packages you can use the following command:

$ conda install -c conda-forge pandas
$ conda install -c r rpy2

7. R packages: randomForest, e1071, xgboost and stringr

$ install.packages('randomForest')
$ install.packages('e1071')
$ install.packages('xgboost')
$ install.packages('stringr')

1. Download and/or install prerequisites described above.

2. Download and unpack the distribution:

$ wget https://github.com/minghuilab/PremPS/archive/v1.0.0.tar.gz
$ tar -zxvf v1.0.0.tar.gz

3. Change to the source directory:

$ cd PremPS-1.0.0

4. Change the path parameters in PremPS.py (line 15-20):

workdir = Your working directory
pathvmd = path for running VMD software  # /usr/local/bin/vmd
pathmkdssp = path for running DSSP software  # /usr/local/bin/mkdssp
pathpsiblast = path for running PSI-BLAST software  # /usr/local/bin/blast/psiblast
pathblastdb = path for blastdb  # /usr/local/bin/blastdb/nr
pathrscript = path for running Rscript  # /usr/local/bin/Rscript

    The FoldX software needs to be installed in the working directory.

$ python PremPS.py -i 2020100417132606935696574

PremPS is only intended to run on linux operating systems.

You will need to have Python 2 (or 3) and R 3.4.0 (or higher) installed.