e-LEA3D

e-LEA3D performs computer-aided drug design by combining molecular fragment-based de novo design, virtual screening, and combinatorial library generation to identify and optimize molecular scaffolds and substituents.

Key Features:

• De Novo Drug Design Tool: Uses a heuristic approach based on genetic algorithms to optimize a user-specified scoring function that incorporates structure- and ligand-based evaluations, fits fragments or fragment combinations to QSAR models or protein binding sites, enables scaffold-hopping, and scans for substituents on user-specified scaffolds.
• Virtual Screening and Filtering Tool: Screens and filters uploaded compound libraries according to predefined criteria to prioritize candidates from large collections.
• Combinatorial Library Design Tool: Generates combinatorial libraries by combining user-drawn scaffolds with reactants sourced from chemical suppliers to produce diverse compound libraries.

Scientific Applications:

• Scaffold discovery and scaffold-hopping: Exploration and identification of novel molecular scaffolds through fragment recombination and scaffold-hopping strategies.
• QSAR- and structure-based ligand optimization: Optimization of ligands against QSAR models or protein binding sites using fragment-based design and scoring functions.
• Substituent scanning: Systematic scanning of substituents on specified scaffolds to expand chemical space and probe structure-activity relationships.
• Library prioritization: Prioritization and filtering of large compound libraries via virtual screening to identify promising candidates.
• Combinatorial library generation: Design of combinatorial libraries from scaffold and reactant combinations for subsequent evaluation and synthesis planning.

Methodology:

Heuristic genetic algorithms optimize user-specified scoring functions that combine structure- and ligand-based evaluations; fragments are fitted to QSAR models or protein binding sites; uploaded libraries undergo virtual screening and filtering; combinatorial libraries are generated by combining user-drawn scaffolds with reactants.