Skip to content

Latest commit

 

History

History
95 lines (80 loc) · 2.85 KB

README.md

File metadata and controls

95 lines (80 loc) · 2.85 KB
Raw

SPRank - A knowledge-based scoring function for RNA-ligand pose prediction and virtual screening

Platform Requirements (Tested)

The following are tested system settings.

For compiling and running /bin/sprank

  • GNU/Linux x86_64 (Ubuntu 22.04.3 LTS kernel 5.15.0-91-generic)
  • GNU Make 4.3
  • gcc/g++ (version 11.4.0)

For running /util/check_atom_order

  • Python 3.11.5
  • NumPy 1.26.0

For running /util/ambertools_prepare_rec and /util/ambertools_prepare_cpd

  • AmberTools22

For running random forest model attached in the Releases

  • scikit-learn 1.5.0
  • Polars 0.20.4
  • NumPy 1.26.0
  • Matplotlib 3.7.2

Installation

Compile and install SPRank

Clone this repository on your local machine and run setup script

git clone https://github.com/Vfold-RNA/SPRank.git ${HOME}/SPRank

compile the code

cd ${HOME}/SPRank && make

put the following environment variable to your .bashrc

echo "export SPRANK_HOME=${HOME}/SPRank" >> ${HOME}/.bashrc

and source it

source ${HOME}/.bashrc

Using SPRank

Check SPRank options

${SPRANK_HOME}/bin/sprank -h

Run SPRank for example cases

cd ${HOME}/SPRank/example && chmod +x ./run_example && ./run_example

The predicted scores will be saved in the corresponding folders as score.dat. By default, the script does not run AmberTools22 to prepare the input receptor and compound. You can remove the comments in ./run_example to prepare the input files using AmberTools22.

SPRank command line arguments

-r <receptor>         # path to target RNA (in amber mol2 format, must contain hydrogens)
-c <target compound>  # path to target compound (in amber mol2 format, must contain hydrogens 
                        and bond table (i.e., "@<TRIPOS>BOND" record))
-p <compound poses>   # path to poses sampled by docking software,
                        to be scored by SPRank (in mol2 format,
                        the order of the heavy atoms should be same as the target compound)

Download data

The training set, pose sets, affinity sets, random forest model, amber atom types, potentials, HIV-1 TAR ensemble and compound library can be downloaded from the Releases or through the following commands:

mkdir -p ${HOME}/SPRank/data/

for name in "checksum.txt" "training-set.tar.gz" "pose-sets.tar.gz" "affinity-sets.tar.gz" "random-forest.tar.gz" "amber-types.tar.gz" "potentials.tar.gz" "HIV-1-TAR.tar.gz"
do
    wget https://github.com/Vfold-RNA/SPRank/releases/download/data/${name} -O ${HOME}/SPRank/data/${name}
done

Check the integrity of the files:

cd ${HOME}/SPRank/data/

sha256sum --check checksum.txt

References

[1] Zhou Y, Jiang YW, Chen SJ. SPRank - A Knowledge-Based Scoring Function for RNA-Ligand Pose Prediction and Virtual Screening. Journal of Chemical Theory and Computation. 2024 Aug. doi: 10.1021/acs.jctc.4c00681.