Spacio-Linear Screening for Ligand-Docking Cavities in Protein Structures: SLAM Algorithm

Area of Science:

• Computational Biology
• Structural Bioinformatics
• Drug Discovery

Background:

• Identifying structurally similar ligand-binding sites aids drug repurposing and understanding protein function.
• Existing structural screening tools often lack sensitivity and scalability.

Purpose of the Study:

• To develop a novel algorithm for discovering novel protein targets for a ligand using a known ligand-binding query protein.
• To present SLAM (Spacio-Linear Alignment of Macromolecules) for detecting local 3D similarities between protein structures.

Main Methods:

• SLAM encodes spatial substructure neighborhoods into linear sequences of annotated atoms.
• Pairwise sequence alignment and distance-correlation scoring identify structural matches.
• Benchmarking performed using the Kahraman-36 dataset against the ProBiS algorithm.

Main Results:

• SLAM outperforms ProBiS in true-positive rate for predicting ligand-docking compatibility.
• Identified potential inhibitors for CRISPR-Cas proteins.
• Predicted novel binding partners for toxic per- and polyfluoroalkyl substances (PFAS).

Conclusions:

• SLAM is a robust, efficient, and flexible tool for structural screening.
• It detects subtle physicochemical compatibilities between protein surfaces.
• SLAM promises to accelerate target discovery in pharmacology and environmental toxicology.