PharmMapper

PharmMapper facilitates in silico drug target identification by pharmacophore mapping of small molecules to receptor-based models.

Key Features:

• Pharmacophore mapping: Maps query small molecules against receptor-based pharmacophore models to predict spatial arrangements of interaction features.
• PharmTargetDB: Uses an in-house database containing over 7,000 receptor-based pharmacophore models derived from TargetBank, BindingDB, DrugBank, and other drug target resources, covering more than 1,500 drug targets.
• Triangle hashing mapping: Employs a triangle hashing mapping method to enable high-throughput matching between molecules and pharmacophore models.
• Full-database screening and ranking: Maps each query against all pharmacophore models, ranks interactions, and outputs the top N best-fitted hits with target annotations and aligned molecular poses.
• Approach versus docking: Predicts feature-based spatial arrangements using pharmacophore models rather than using traditional molecular docking methods.
• Throughput: Database-wide screening typically completes in about one hour.
• Validation: Validated in retrospective tests, for example identifying correct tamoxifen targets among the top 300 pharmacophore candidates.

Scientific Applications:

• Drug target identification: Identifying potential protein targets for drugs, natural products, and novel compounds with unknown binding targets.
• High-throughput virtual screening: Screening candidate molecules against a large collection of receptor-based pharmacophore models.
• Target annotation and pose prediction: Providing ranked target annotations and aligned molecular poses for downstream analysis.

Methodology:

Maps query molecules against PharmTargetDB’s >7,000 receptor-based pharmacophore models using a triangle hashing mapping method to predict spatial arrangements of interaction features, ranks the fits, and returns the top N best-fitted hits with target annotations and aligned molecular poses; the models were derived from TargetBank, BindingDB, DrugBank and other drug target resources and cover >1,500 drug targets.