OpenPharmaco Manual ENG
OpenPharmaco powers PharmacoNet's graphical user interface for fully automated protein-based pharmacophore modeling and ultra-fast virtual screening.
Development and maintenance: Seonghwan Seo (Laboratory of Prof. Woo Yeon Kim, Department of Chemistry, KAIST)
If you use OpenPharmaco in your work, please cite:
Seo, Seonghwan, and Woo Youn Kim. "PharmacoNet: deep learning-guided pharmacophore modeling for ultra-large-scale virtual screening." Chemical Science (2024).
Paper: https://doi.org/10.1039/D4SC04854G
Source Code: https://github.com/SeonghwanSeo/OpenPharmaco
Tested Environment: Microsoft Window, Mac OS X (Apple Silicon)
Last modified: 2024.11.24 (Version 1.0.0)
Download the source code from Github and install the required libraries through the Anaconda virtual environment.
# Install Project
git clone https://github.com/SeonghwanSeo/OpenPharmaco
cd OpenPharmaco
# Install conda environment
conda env create -f environment.yml
conda activate openph
pip install .User can now run OpenPharmaco with the openph or openpharmaco command as shown below.
conda activate openph # (re-activate to load alias)
openphThe first time user run OpenPharmaco, it goes through an initial setup process and downloads the necessary files. (139 MB)
Once all downloads have finished, subsequent runs will take about seconds of loading time.
OpenPharmaco downloads protein files from RCSB, automatically parses them, and recognizes binding sites.
Enter the PDB code in Enter RCSB PDB Code.
If user wants to use a custom protein file, manually enter the protein file (.pdb) via Open Protein File and the binding ligand (.mol2, .sdf) via Open Ligand File to recognize the binding site.
Tap the button (▸) left to Protein in the explorer to visualize the structure, including chains, sticks, and lines.
Select the desired binding site ligand from explorer, and click the Modeling to perform protein-based pharmacophore modeling.
The calculated pharmacopore model can be saved by clicking Save Model File (.pm extension)
User can load the saved pharmacopore models by clicking Open Model File, or drag and drop the file into the program window.
Before clicking Open Model File or shortcut, user needs to initialize the session by Clear.
To reset the protein or pharmacophore model, click Clear or press Ctrl-W/Cmd-W.
With the pharmacopore model open, user can perform virtual screening by clicking the Screening.
Click the Library to select a folder, or drag and drop it into the screening window.
Then all sdf and mol2 files within the folder and its subfolders will be loaded.
(Each file can contain one or more conformers).
For example, if user select the library/ folder below, the files a.mol2, b.mol2, c.mol2, d.sdf, d.mol2, and e.sdf will all be loaded.
├── library/
├── a.mol2
├── library1/
│ ├── b.mol2
│ └── library1_1/
│ ├── c.mol2
│ └── d.sdf
└── library2/
├── d.mol2
└── e.sdfBelow is a screenshot when opening an example library of 1,000 compounds containing 8 conformers. (Example library: https://drive.google.com/file/d/1XCMv97WpfgEccR4xXTGep_PCMMJq3f7t/view?usp=share_link)
User can change the file parsing rule by clicking Advanced and choose between File Name or File Path. (Default: File Name.)
For example, d.sdf and d.mol2 above are indistinguishable because they have the same filename, but they are distinguishable with their file paths.
Below is the screenshot when File Path is selected.
PharmacoNet, a built-in program in OpenPharmaco, performs screening through parameters assigned to 7 pharmacophoric feature type. OpenPharmaco allows user to change the weighting for each type.
User can change the value under Parameter Setting via mouse or keyboard, and it will be colored blue if it is set higher than the default, or red if it is set lower.
Clicking Reset All resets the parameters to the default setting.
Once everything is set up, set the number of CPU cores for multi-processing by changing the slide bar or number next to CPUs.
OpenPharmaco automatically recognizes the maximum number of CPUs in system.
Finally, click Run to start virtual screening.
After screening, the molecules with the highest scores are listed first.
Click Save to export the screening result.
The key is determined by the user setting: File Name or File Path.
To run screening with different libraries or parameter settings, click Clear to initialize the session.
OpenPharmaco was developed utilizing PyTorch, NumPy, Numba, BioPython, OpenBabel, Open-Source PyMol, and PyQt5.
Paper: https://arxiv.org/abs/2310.00681
If you use OpenPharmaco in your work, please cite:
Seo, S., & Kim, W. Y. (2023, December). PharmacoNet: Accelerating Large-Scale Virtual Screening by Deep Pharmacophore Modeling. In NeurIPS 2023 Workshop on New Frontiers of AI for Drug Discovery and Development.
MIT License
Copyright (c) 2023 Seonghwan Seo
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.