ChemBioServer 2.0

ChemBioServer 2.0 facilitates analysis, filtering, clustering, and structural-networking of chemical compound libraries to support lead identification, lead optimization, and drug repurposing.

Key Features:

• Compound visualization: Visualizes compounds with associated physicochemical properties and predicted toxicity profiles.
• Property-based filtering: Filters compounds according to specified physicochemical and toxicity criteria.
• Perfect-match substructure search: Performs perfect-match substructure searching for substructure-based mining and lead optimization.
• Re-ranking of virtual screening results: Re-ranks virtual screening outputs to prioritize compounds with high affinity for a target while minimizing interactions with other family members.
• Clustering by physicochemical properties: Clusters compounds based on physicochemical properties and reports representative compounds for each cluster.
• Structural similarity network construction and analysis: Constructs structural similarity networks and computes network analysis metrics for relationship analysis and visualization.
• Integration of compound sets for repurposing: Merges query and reference compound sets into a single structural similarity network to reveal repurposing opportunities via transitive similarities.
• Exclusion of unwanted compounds by similarity: Removes compounds from networks based on similarity to unwanted substances, including previously failed drugs.
• Custom compound-mining pipelines: Supports construction of custom compound-mining pipelines for tailored analyses.

Scientific Applications:

• Lead identification: Uses filtering, clustering, and visualization to identify candidate lead compounds from libraries.
• Lead optimization: Applies perfect-match substructure search and property filters to support lead optimization efforts.
• Virtual screening prioritization: Re-ranks virtual screening results to enhance target selectivity among screened compounds.
• Drug repurposing via network transitivity: Integrates compound sets into structural similarity networks to identify repurposing opportunities through transitive similarities.
• Exclusion of undesirable compounds: Removes compounds similar to previously failed drugs to refine candidate sets.
• Comparative selection across chemical space: Provides representative compounds per cluster to aid comparative analysis and selection.

Methodology:

Implements perfect-match substructure searching, property-based filtering, clustering based on physicochemical properties, structural similarity network construction and network analysis metrics, re-ranking of virtual screening results, integration of compound sets into single networks, and removal of compounds based on pairwise similarity.