Structure-Based Drug Design (SBDD)

Method

Structure-Based Drug Design is a method of drug discovery that relies on the three-dimensional structure of a target protein obtained through techniques such as X-ray crystallography, NMR spectroscopy, or cryo-electron microscopy. By understanding the 3D structure, researchers can design molecules that fit precisely (ligands) into the protein's active or binding sites. If a structure is not available, homology modeling or deep learning protein prediction methods can be used. Otherwise, ligand-based drug design might be more appropriate.

Importance in Computational Drug Discovery:

Precision Targeting: SBDD enables the design of ligands that specifically fit the protein's binding site, potentially leading to higher efficacy and fewer off-target effects.
Optimization: Researchers can iteratively modify and optimize ligands based on their interactions with the target protein.
Efficiency: Computational tools can rapidly screen large libraries of compounds, identifying those with the best binding affinities and other desirable properties.
Mechanistic Insights: SBDD provides insights into the molecular mechanisms of drug action, helping in understanding how drugs interact with their targets at the atomic level.

‍

Key Tools

DecompDiff: Diffusion Models with Decomposed Priors for Structure-Based Drug Design

Publication Date: 2024-02-26
DOI: 10.48550/arXiv.2403.07902
Summary: This paper introduces a new diffusion model, DecompDiff, which decomposes ligand molecules into arms and scaffold to improve the generation of high-affinity molecules.

Accelerated Discovery of Macrocyclic CDK2 Inhibitor QR-6401 by Generative Models and Structure-Based Drug Design

Publication Date: 2023-02-08
DOI: 10.1021/acsmedchemlett.2c00515
Summary: The discovery of a potent macrocyclic CDK2 inhibitor accelerated by generative models and SBDD, demonstrating robust antitumor efficacy.

Drug Discovery for Mycobacterium tuberculosis Using Structure-Based Computer-Aided Drug Design Approach

Publication Date: 2021-12-01
DOI: 10.3390/ijms222413259
Summary: This review highlights computational techniques in designing new anti-tubercular drugs, emphasizing the integration of machine learning, AI, and quantum computing.

Learning Subpocket Prototypes for Generalizable Structure-based Drug Design

Publication Date: 2023-05-22
DOI: 10.48550/arXiv.2305.13997
Summary: A novel method, DrugGPS, is proposed for generalizable SBDD, outperforming baselines in generating high-affinity drug candidates.

Hybrid Quantum Mechanics/Molecular Mechanics (QM/MM) Simulation: A Tool for Structure-based Drug Design and Discovery

Publication Date: 2021-10-07
DOI: 10.2174/1389557521666211007115250
Summary: This review explores the use of hybrid QM/MM simulations for ligand and structure-based methods in drug design.

Molecular Docking and Structure-Based Drug Design Strategies

Publication Date: 2015-07-01
DOI: 10.3390/molecules200713384
Summary: Examines current molecular docking strategies in drug discovery, exploring advances and the role of integrating structure- and ligand-based methods.

Advances in Computational Structure-Based Drug Design and Application in Drug Discovery

Publication Date: 2015-11-25
DOI: 10.2174/1568026615666150825142002
Summary: An overview of key mechanisms, computer programs, and successful cases in SBDD.

Structure-Based Macrocycle Design in Small-Molecule Drug Discovery

Publication Date: 2019-03-22
DOI: 10.1021/acs.jmedchem.8b01985
Summary: Reviews structure-based design of synthetic macrocycles and the initial evaluation of molecules as candidates for macrocyclization.

Structure-based drug design with geometric deep learning

Publication Date: 2022-10-19
DOI: 10.48550/arXiv.2210.11250
Summary: Provides an overview of geometric deep learning applications in SBDD, emphasizing molecular property prediction, ligand binding site and pose prediction, and structure-based de novo molecular design.

‍

Literature

1. AutoDock Vina:

◦ A widely-used molecular docking software that predicts the preferred binding positions of a ligand to a target protein.

2. Schrödinger Suite:

◦ Includes tools like Glide for docking, Prime for protein structure prediction and refinement, and Maestro for visualization and analysis.

3. GROMACS:

◦ A powerful molecular dynamics simulation software used to study the dynamic behavior of protein-ligand complexes.

4. DeepOrigin Tools:

◦ Docking: For simulating the docking of ligands into protein binding sites.

◦ PocketFinder: For identifying potential binding pockets on proteins.

◦ BindingDBToolExecutor: For retrieving experimental binding data from databases like BindingDB.

5. PyMOL:

◦ A molecular visualization system used to view and analyze protein structures, aiding in the understanding of protein-ligand interactions.

6. Rosetta:

◦ A suite of software for protein modeling and design, including tools for protein-ligand docking and structure prediction.

Explore more materials, our research and analyses of AI and the drug discovery industry below.

Structure-Based Drug Design (SBDD)

Key Tools

Literature

Deep Origin Resources

Scientific Poster

•

Drug Discovery

Discovering Novel Synthetic Lethal Pairs With Large Scale Cellular Simulations

Scientific Poster

•

Drug Discovery

Drug Discovery Challenges with a Multiscale Molecular Modeling Pipeline

Blog

•

Drug Discovery

Finally, a Useful AI Assistant for Drug Discovery: Meet Balto

Discovering Novel Synthetic Lethal Pairs  With Large Scale Cellular Simulations