SwissTargetPrediction

SwissTargetPrediction predicts molecular targets of bioactive small molecules to identify protein and macromolecular binders and elucidate mechanisms of action for compounds such as drugs and metabolites.

Key Features:

• 2D and 3D similarity measures: Employs both 2D and 3D similarity metrics to compare query molecules to reference compounds.
• Reference library size: Compares queries against a library of over 280,000 compounds known to be active on more than 2,000 targets.
• Cross-species coverage: Includes target data spanning five different organisms.
• Homology mapping: Supports mapping predictions by homology within and between species, including close paralogs and orthologs.
• Primary and secondary target prediction: Predicts potential new targets for uncharacterized molecules and secondary (off-)targets for characterized molecules.
• Protein–small molecule interaction data: Integrates large datasets of protein–small molecule interactions using a knowledge-based approach.

Scientific Applications:

• Mechanism elucidation: Identifies candidate protein targets to help clarify molecular mechanisms underlying biological activity.
• Off-target and side-effect prediction: Reveals potential off-targets to support assessment of side effects and cross-reactivity.
• Target identification for drug research: Supports target identification and hypothesis generation in drug discovery and development.
• Cross-species target inference: Enables prediction of targets across species by leveraging homology relationships.

Methodology:

Computational prediction is based on 2D and 3D similarity measures comparing query molecules to a reference set of >280,000 compounds active on >2,000 targets across five organisms, incorporates homology mapping (paralogs and orthologs), and integrates large protein–small molecule interaction datasets in a knowledge-based framework.