Repository for code published in "Streamlining Computational Fragment-Based Drug Discovery through Evolutionary Optimization Informed by Ligand-Based Virtual Prescreening."
The code implements a computational drug design platform that builds on our group's recently published computational fragment-based drug discovery (FBDD) method called fragment databases from screened ligand drug discovery (FDSL-DD). FDSL-DD uses in silico screening to identify ligands from a vast library, fragmenting them while attaching specific attributes based on predicted binding affinity and interaction with the target subdomain. In this paper, we further propose a two-stage optimization method that utilizes the information from prescreening to optimize computational ligand synthesis. We hypothesize that using prescreening information for optimization shrinks the search space and focuses on promising regions, thereby improving the optimization for candidate ligands. The first optimization stage assembles these fragments into larger compounds using genetic algorithms, followed by a second stage of iterative refinement to produce compounds with enhanced bioactivity.
| Folder | Description |
|---|---|
| genetic | Code to run genetic algorithm |
| iterative | Code to run iterative fine tuning algorithm |
| inputs | Data used as inputs for genetic/iterative algorithms |
| results | Contains generated ligands and associated scores |
| autodock_config_files | Files to configure Autodock VINA runs |
| fragmentation_pipeline | Used to generate inputs, from previous work |
Environment packages can be found in environment.yml. In addition, an install of rdkit, vina=1.2.3, and openmpi are required to run the scripts. Singularity is also necessary to run PLIP, which has documentation that can be found at https://github.com/pharmai/plip?tab=readme-ov-file. Version 2.2.0 was used for this project.
The genetic algorithm is ran using the genetic.py script in the genetic folder. The genetic.sh script in the genetic folder provides a command line template as well as SLURM parameters to run this script. To merge final results, use the mergeResults.py script to generate a single csv of ligands and associated binding affinities with QED scores. Inputs to this algorithm for the proteins presented in the paper are found in the inputs folder.
The iterative algorithm is ran on the results from the genetic algorithm. The iterative.py script within the iterative folder coordinates this. The iterative.sh script contains a command line template as well as SLURM parameters to run this script. To run PLIP, be sure to change the PLIP .smg file to the name of the installed singularity file. Use mergeResults.py script to generate single csv of results.
Note: Code assumes it is being run on a SLURM cluster, and requires certain environmental variables like $TMPDIR, $SLURM_ARRAY_TASK_ID, and $SLURM_JOB_ID which may not be present on all systems. These are typically used for file handling, and may be adapted based on individual use cases. For greater compatibility, a more portable version is under development.
Individuals interested in generating novel fragments based off additional ligand libraries or target proteins should view the fragmentation_pipeline folder, which is described at https://doi.org/10.1016/j.jmgm.2023.108669. Note this code is separate from the project described in this paper and has it's own README.txt file. A dataset of prescreened ligands from Autodock VINA is needed to generate fragments using this methodology.
Feel free to reach out to the authors with any questions.
The methodology is fully described, along with experimental results and comparisons to DeepFrag and AutoGrow, in the following publication:
Chandraghatgi R, Ji HF, Rosen GL, Sokhansanj BA. Streamlining Computational Fragment-Based Drug Discovery through Evolutionary Optimization Informed by Ligand-Based Virtual Prescreening. J Chem Inf Model. Published online May 2, 2024. doi:10.1021/acs.jcim.4c00234
Shield:
This work (including any and all source code and accompanying documentation in this repository) is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License.
