This website is free and open to all users and there is no login requirement.
 

InterEvData Code



How to use the provided Singularity Image?



All the tools that we present below are provided and packaged into an easy-to-use Singularity image an simply require that singularity be installed on your computer
or cluster. See FAQ question 3 for troubleshooting and tips on how to install Singularity.



1. Running the complete docking pipeline
To perform the whole docking pipeline (sampling+scoring), you can simply run the docking_pipeline.py script within the Singularity image. The script can be executed as so: 
singularity exec interevdata_tools.sif docking_pipeline.py [options]
The basic pipeline:

The basic pipeline (Frodock sampling, Soap-PP+InterEvScore scoring & consensus selection) can be run on 2 single-chain protein structures (receptor.pdb and ligand.pdb) with the following command: 
singularity exec interevdata_tools.sif docking_pipeline.py -R receptor.pdb \
        -L ligand.pdb -r receptor_coMSA.fasta -l ligand_coMSA.fasta
N.B. Please note that the pipeline doesn't generate the co-Multiple Sequence Alignments (co-MSAs) but you can add them as inputs to the pipeline with options -r and -l in order to fully benefit from the co-evolutionary-based score InterEvScore. For PPI4DOCK cases, the co-MSAs are provided in the compressed folder (see InterEvData home), otherwise, they can be generated using InterEvolAlign for example (see InterEvolAlign))

Using the explicit homology feature:

To run the explicit homology feature (or to run scoring with Rosetta's ISC), you will need a version of the Rosetta package already installed (see FAQ question 3 for the installation link) and you will have to link the folders to the singularity image at execution. The --runExplicit and --runISC options will enable you to activate the homology modelling feature and the ISC scoring respectively: 
singularity exec -B your_rosetta:/usr/local/rosetta_dir interevdata_tools.sif \
        env LD_LIBRARY_PATH=/usr/local/rosetta_dir/$LIB_PATH docking_pipeline.py \
        --runExplicit --runISC [options]
your_rosetta should be the path to your Rosetta directory (for example rosetta_2017.08.59291_bundle). LIB_PATH should correspond to Rosetta's library paths within the Singularity image. The value of LIB_PATH might need adjusting according to your Rosetta version but must be given in reference to the path inside the image. For example: 
LIB_PATH=main/source/build/external/release/linux/4.9/64/x86/gcc/6.3/default/
Running the pipeline in parallel:

The docking pipeline can be run in parallel by specifying options -n (number of nodes) and -p (number of cpus per node) provided that the indicated number of resources (nodes*cpus) is available: 
singularity exec interevdata_tools.sif docking_pipeline.py -n 2 -p 8 [options]
The expected output is detailed and explained in FAQ question 4.



2. Generating all the decoy structures
You might want to use the Singularity image in order to generate the query and homolog decoy structures so that you can score them with your own scoring function. Once you've executed the docking pipeline, you can easily generate the decoys with the generate_decoys.py script that is within the Singularity image: 
singularity exec interevdata_tools.sif generate_decoys.py -r list_receptors.txt \
        -l list_ligands.txt -d dockclust.dat -o output_directory [options]
list_receptors.txt and list_ligands.txt should be files containing a list of paths for the query and homolog modelled unbound structures and dockclust.dat the frodock sampling output (translation and rotation coordinates). Typically, use can use the dockclust.dat and the lists outputed by the docking pipeline described above (rosetta_homologs_partner_{a,b}/homolog_partner_{a,b}_list.txt, see FAQ question 4).

3. Scoring your own decoys with Soap-pp and InterEvScore
If you want to score your own decoy structures with the Soap-PP and InterEvScore scoring functions provided in the Singularity image, you can do so by executing the score_decoys.py script that is within the Singularity image: 
singularity exec interevdata_tools.sif score_decoys.py -f list_decoys.txt \
        -c chains1:chains2 -o output_scores.txt -r receptor_coMSA.fasta \
        -l ligand_coMSA.fasta [options]
list_decoys.txt should be a file containing a list of paths to your decoy structures. N.B. Please note that to fully benefit from the co-evolutionary-based score InterEvScore, you should add the co-Multiple Sequence Alignments to the script with options -r and -l.For PPI4DOCK cases, the co-MSAs are provided in the compressed folder (see InterEvData home), otherwise, they can be generated using InterEvolAlign for example (see InterEvolAlign))



 Help
More information on execution can be displayed with Singularity's help (version 2) or run-help (version 3) option or by running one of the three python scripts above with the --help option. 
singularity help interevdata_tools.sif
singularity exec interevdata_tools.sif docking_pipeline.py --help
Depending on your version of Singularity, you might also have to explicitly add your current working directory at the execution of the Singularity image with -B $PWD. For more troubleshooting advice, see FAQ question 3.


How to cite us:
Chloé Quignot, Pierre Granger, Pablo Chacón, Raphaël Guerois and Jessica Andreani: Atomic-level evolutionary information improves protein-protein interface scoring (BioRxiv doi:10.1101/2020.10.26.355073)