V-Dock: Fast Generation of Novel Drug-like Molecules Using Machine-Learning-Based Docking Score and Molecular Optimization

Abstract

We propose a computational workflow to design novel drug-like molecules by combining the global optimization of molecular properties and protein-ligand docking with machine learning. However, most existing methods depend heavily on experimental data, and many targets do not have sufficient data to train reliable activity prediction models. To overcome this limitation, protein-ligand docking calculations must be performed using the limited data available. Such docking calculations during molecular generation require considerable computational time, preventing extensive exploration of the chemical space. To address this problem, we trained a machine-learning-based model that predicted the docking energy using SMILES to accelerate the molecular generation process. Docking scores could be accurately predicted using only a SMILES string. We combined this docking score prediction model with the global molecular property optimization approach, MolFinder, to find novel molecules exhibiting the desired properties with high values of predicted docking scores. We named this design approach V-dock. Using V-dock, we efficiently generated many novel molecules with high docking scores for a target protein, a similarity to the reference molecule, and desirable drug-like and bespoke properties, such as QED. The predicted docking scores of the generated molecules were verified by correlating them with the actual docking scores.

Optimizing interactions to protein binding sites by integrating docking-scoring strategies into generative AI methods

“…Here, ligand structures are directly generated within the binding site and scored using refinement docking of this initial molecule, which appears to be much faster than standard docking. Moreover, a prediction model for docking scores from SMILES as reward function for molecular design was implemented in the program V-dock ( Choi and LeeV-Dock, 2021 ). Finally, in a retrospective design study for the GPCR DDR2, Glide and its Glide-SP score were directly integrated into the REINVENT generative approach ( Thomas et al, 2021 ).…”

Link-INVENT: generative linker design with reinforcement learning

“…(1) Linker effective length = [4,6], [7,9], [10,12], or [13,15]: this component enforces linkers to possess an effective length within the specied intervals. See ESI S15 † for the Scoring Function transformation.…”

“…The curve in (c) shows the average score achieved by the batch of molecules sampled at a given epoch and the upper and lower bounds of the shaded region represent the maximum and minimum scores, respectively. (a) Experiment that fixes physico-chemical properties and tasks Link-INVENT with generating linkers with an effective length within the specified intervals: [4, 6],[7,9],[10,12], and[13,15]. The baseline experiment does not enforce the linker length.…”

Parallel tempered genetic algorithm guided by deep neural networks for inverse molecular design

“…Better than imitating molecular docking with a classier or using ML-based surrogate models, [75][76][77] to approach the complexity of real-life drug discovery tasks, actual molecular docking has been proposed to benchmark generative models. 78 Hence, we tackled the tasks introduced by Cieplinski et al, 78 i.e., minimize the docking scores obtained from actual docking calculations to the protein targets 5HT1B, 5HT2B, ACM2 and CYP2D6.…”