GRADE

This reposatory includes all scripts necessary to calculate the GRAil-based DEscriptor (GRADE) and the eXtended GRAil-based DEscriptor (X-GRADE) and the scripts that are necessary to reproduce the results of the paper "GRADE and X-GRADE: Unveiling novel Protein-Ligand Interaction Fingerprints based on GRAIL-Scores"

Install

To reproduce the results of the paper, one has to install a package called PHANTOMDRAGON (PHarmacophore bAsed scoriNg funcTiOn iMplementations using DRug interAction Grail scOre calculatioNs). For your convinience, the package is available right here in this repository. To install the package, we recommend using a conda or mamba environment:

conda create -n GRADE python==3.10

This should give you a blank environment with only python and its environments installed. For this package python >3.7 and <3.11 is reqired. Then activate the environment with:

conda activate GRADE

Then install the package with:

pip install .

Of course one has to be in the GRADE directory for this to work. This should also install all dependencies that you need.

GRADE/X-GRADE generation

Depending on your data structure, run one of the following commands to generate GRADE or X-GRADE:

calc_descr_pdb_bind.py [-h] -d COMPLEX_DATA_DIR -o OUT_CSV_FILE [-c] [-x]

Calculates GRADE/X-GRADE for a set of input ligand-protein complexes. The Files have to be organized in PDBbind manner.

Option	Description	Required	Default
-h or --help	Show this help message and exit	No	N/A
-d COMPLEX_DATA_DIR	The directory containing the ligand-protein complexes to process, organized in PDBBind manner	Yes	N/A
-o OUT_CSV_FILE	The path of the output CSV-file containing the descriptor values calculated for each input complex	Yes	N/A
-c	Change protonation of acidic/basic groups to a state likely at pH7	No	false
-x	Calculate extended GRAIL descriptor with subdivided HBA/HBD feature types	No	false

calc_descr_PL_REX.py [-h] -d COMPLEX_DATA_DIR -o OUT_CSV_FILE [-c] [-x]

Calculates GRADE/X-GRADE for a set of input ligand-protein complexes. The Files have to be organized in PL-REX manner.

Option	Description	Required	Default
-h or --help	Show this help message and exit	No	N/A
-d COMPLEX_DATA_DIR	The directory containing the ligand-protein complexes to process, organized in PL-REX manner	Yes	N/A
-o OUT_CSV_FILE	The path of the output CSV-file containing the descriptor values calculated for each input complex	Yes	N/A
-c	Change protonation of acidic/basic groups to a state likely at pH7	No	false
-x	Calculate extended GRAIL descriptor with subdivided HBA/HBD feature types	No	false

calc_descr_pdb_ligands.py [-h] -p PDB_FILE -l LIG_FILE -o OUT_CSV_FILE [-x] [-c]

Calculates GRADE/X-GRADE for a PDB-file and set of input ligands.

Option	Description	Required	Default
-h or --help	Show this help message and exit	No	N/A
-p PDB_FILE	The receptor PDB-file	Yes	N/A
-l LIG_FILE	The file providing the ligands	Yes	N/A
-o OUT_CSV_FILE	The path of the output CSV-file containing the descriptors calculated for each input ligand	Yes	N/A
-x	Calculate extended GRAIL descriptor with subdivided HBA/HBD feature types	No	false
-c	Change protonation of acidic/basic groups to a state likely at pH7	No	false

Repoducing the results

You can use one of the scripts in the scripts directory to reproduce the results.

The scripts are (in alphabetical order):

Script Name	Description	Type
3DQSAR_GRADE.ipynb	Performs 3D QSAR analysis using parts of GRADE and X-GRADE.	Jupyter Notebook
calc_descr_pdb_bind.py	Calculates GRADE/X-GRADE for a set of input ligand-protein complexes. The Files have to be organized in PDBbind manner. (see above)	Python Script
calc_descr_pdb_ligands.py	Calculates GRADE/X-GRADE for a PDB-file and set of input ligands. (see above)	Python Script
calc_descr_PL-REX.py	Calculates GRADE/X-GRADE for a set of input ligand-protein complexes. The Files have to be organized in PL-REX manner. (see above)	Python Script
create_core_set_data.py	Creates the PDBbind core dataset for model evaluation.	Python Script
create_general_set_data.py	Creates the PDBbind general set for model testing.	Python Script
create_refined_set_data.py	Creates the PDBbind refined set for modle training.	Python Script
extract_data.ipynb	Extracts data from various sources and formats it for analysis.	Jupyter Notebook
pred_classes.py	Estimates binding affinities for different EC numbers.	Python Script
pred_PL-REX.py	Estimates binding affinities for the PL-REX Dataset.	Python Script
pred_test_set.py	Evaluates estimations on a test set.	Python Script
pred_validation_set.py	Evaluates estimations on a validation set.	Python Script
test_time.py	Measures and tests the runtime of the binding affinity estimations.	Python Script
UMAP_generation.ipynb	Generates UMAP visualizations for data analysis. Also includes t-SNE.	Jupyter Notebook

Credits

CDPKit and the GRADE and X-GRADE implementation was done by Thomas Seidel (https://github.com/seidelt)

PHANTOMDRAGON (https://github.com/molinfo-vienna/phantomdragon), all testing and this Reposatory was created by Christian Fellinger (https://github.com/Dragon3221)

This package was created with the help of Cookiecutter_ and the audreyr/cookiecutter-pypackage_ project template.

.. _Cookiecutter: https://github.com/audreyr/cookiecutter .. _audreyr/cookiecutter-pypackage: https://github.com/audreyr/cookiecutter-pypackage