FragmentStore

FragmentStore provides a searchable database and analysis platform for 35,000 molecular fragments derived from 13,000 metabolites, 16,000 drugs, and 2,200 toxic compounds to support fragment-level analyses of biosynthesis and degradation pathways, protein–ligand interactions, toxicity prediction, and small-molecule library generation.

Key Features:

• Database composition: Contains 35,000 distinct building blocks derived from 13,000 metabolites, 16,000 drugs, and 2,200 toxic compounds.
• Fragment generation and relevance: Systematically generates molecular fragments used to analyze biosynthesis, degradation, and links to side-effects and toxicity profiles.
• Search capabilities: Enables filtering and retrieval by 2D structure, molecular weight, and rotatable bonds.
• Amino acid preferences: Calculates amino acid preferences at fragment binding sites to inform protein–ligand interaction predictions.
• Fragment classification: Classifies fragments based on the enzyme classification class of interacting enzymes.
• Small-molecule library generation: Provides a fragment-assembler to generate small molecule libraries from fragments.
• Hypothesis generation: Identifies common fragments across metabolite, drug, and toxic compound classes to suggest shared targets or potential off-target effects and overlaps with toxic compounds.

Scientific Applications:

• Pharmacology: Supports fragment-based drug design and identification of shared fragment signatures between drugs that may indicate similar targets.
• Toxicology: Aids toxicity prediction by linking fragments to known toxic compounds and side-effect profiles.
• Biochemistry and metabolism: Facilitates elucidation of metabolic pathways through fragment analysis of biosynthesis and degradation.
• Structural biology: Informs protein–ligand interaction studies via amino acid preference calculations at fragment binding sites.
• Small-molecule discovery: Enables construction of small-molecule libraries from biologically relevant fragments for screening and lead generation.

Methodology:

Systematic generation of molecular fragments, search by 2D structure, molecular weight and rotatable bonds, calculation of amino acid preferences at fragment binding sites, classification of fragments by enzyme classification class, fragment assembly into small-molecule libraries, and identification of common fragments across molecule classes.