Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2025
|
| Online Access: | https://doi.org/10.5281/zenodo.17498468 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <p><span>The core goal of medicinal chemistry is to discover and optimize drug molecules with high activity, low toxicity, good synthesizability, and target specificity. Traditional drug development cycles are long and costly, but the introduction of artificial intelligence (AI) offers a novel, data-driven approach to drug molecule optimization. This paper proposes a novel AI-assisted medicinal chemistry optimization algorithm, the Molecular Adaptive Heuristic Reinforcement Optimization (MAHRO). The algorithm's core innovations include the introduction of a dynamic heuristic adaptation mechanism (AHM) to adaptively adjust optimization objective weights, reward decomposition and gradient fusion (RDGF) to update multi-objective optimization strategies, and the utilization of a structure-aware generative policy (SAGP) for efficient molecular engineering. This paper systematically describes the MAHRO algorithm's molecular representation, heuristic scoring, reward design, reinforcement learning strategy, and adaptive mechanism. The algorithm's principles are detailed in a complete mathematical formulation, aiming to provide a theoretical foundation and methodological reference for AI-assisted medicinal chemistry.</span></p>