Home  Contact Us  
  • For more information about
    Tripos products and services
    please contact us:
    800-323-2960

  • Tel: +1-314-647-1099
  • Fax: +1-314-647-9241
Home > Products > SYBYL-X > Ligand-Based Design > GALAHAD

GALAHAD

Rapid, High Quality Pharmacophoric Perception and Molecular Alignments

Overview

GALAHAD® aligns a set of molecules that share a common mode of biological activity, and develops a corresponding pharmacophore hypothesis. Using a sophisticated new genetic algorithm and a multi-objective scoring function, GALAHAD takes into account energetics, steric similarity, and pharmacophoric overlap, while accommodating conformational flexibility, ambiguous stereochemistry, alternative ring configurations, multiple partial match constraints, and alternative feature mappings among molecules. Pharmacophore models are returned as hypermolecules, which contain information from every molecule in the training set, as well as a 3D search query that can be used to probe databases for new structures that match the model. New target molecules that were not included in the training set can be fit to the model, yielding scores for energy as well as steric and pharmacophoric similarity that relate directly to ligand affinity.

GALAHAD allows researchers to automatically develop pharmacophore hypotheses and structural alignments from a set of molecules that bind at a common site. No prior knowledge of pharmacophore elements, constraints, or molecular alignment is required, making it ideal for exploring new targets and new modes of action.

GALAHAD uses a sophisticated new genetic algorithm (GA) that defines each molecule as a core structure plus a set of torsions. To overcome limitations in existing pharmacophore tools, GALAHAD's genetic algorithm was developed on real-world data sets.

GALAHAD Brochure (243k)

GALAHAD model derived from four cyclin-dependent kinase (CDK2) inhibitors (left) vs. the overlay based on the corresponding X-ray crystal structures (right).

 

 

Key Benefits

  • Pareto multi-objective optimization is used to simultaneously balance steric, pharmacophoric, and energy information to build the most valuable hypermolecule models required, so models are unbiased
  • Run time scales linearly with the number of ligands, unlike other methods
  • Partial match and partial coverage models can be created in a timely manner