D3Similarity

D3Similarity predicts protein targets and performs virtual screening by evaluating two-dimensional (2D) and three-dimensional (3D) molecular similarity to support ligand-based drug discovery for SARS, MERS, and SARS-CoV-2.

Key Features:

• Ligand-Based Approach: Uses a ligand-centric strategy to mitigate false negatives arising from varied binding-pocket conformations and incomplete protein-target databases.
• 2D and 3D Molecular Similarity: Employs both two-dimensional and three-dimensional structural similarity evaluations for compound comparison.
• Database Composition: Compares submitted molecules against a curated database of bioactive molecules known to interact with coronaviruses (SARS, MERS, SARS-CoV-2), including compounds with and without documented target or mechanism information.
• TargetPrediction: Predicts potential target proteins for queried molecules based on similarity to known bioactive compounds.
• VirtualScreening: Screens and ranks active compounds against specified target proteins using similarity-derived scores.
• Similarity-Based Ranking: Ranks candidate compounds by structural similarity scores derived from 2D and 3D comparisons.
• Complementary to Structure-Based Docking: Serves as a complementary method to protein structure-based docking approaches such as D3Docking.

Scientific Applications:

• Drug Discovery: Identifies candidate therapeutic compounds for SARS, MERS, and SARS-CoV-2 by similarity to known bioactive molecules.
• Target Prediction: Infers likely protein targets for novel or active compounds to aid mechanism-of-action studies.
• Complementary Screening Strategy: Augments protein structure-based docking workflows by providing ligand-based evidence for target selection and prioritization.

Methodology:

Computationally compares submitted molecules to a curated database of coronavirus-related bioactive compounds using 2D and 3D molecular similarity and ranks compounds by similarity scores to inform target prediction and virtual screening.