AI-Predicted Viral Protein Complex Explorer
Investigate 3D quaternary interfaces, AI confidence scores (ipTM & pLDDT), receptor binding affinities, and in silico mutation stability across high-priority viral outbreak complexes.
3D Quaternary Structure View
Click & drag to rotate • Scroll to zoom • Right-click to panOpen Viral Structural Genomics
In international collaborations led by AI research teams, open access to over 2,800+ computationally folded viral protein complexes equips virologists and epidemiologists with atomic-level blueprints of pathogen-host interaction sites long before wet-lab cryo-EM or crystallography assays can be crystallized.
This interactive workbench visualizes quaternary docking poses, inter-chain buried solvent accessible surface area (SASA), interface hydrogen bond networks, and simulated Gibbs free energy shifts (ΔΔG) induced by viral escape mutations.
Frequently Asked Questions
What is the interface predicted TM-score (ipTM)?
ipTM measures the accuracy of predicted inter-chain interfaces in a multimer complex on a scale from 0 to 1. Values above 0.80 represent high confidence relative orientations suitable for structure-guided antiviral discovery.
How is the binding energy (ΔΔG) computed?
Free energy shifts are modeled via physical empirical potential functions evaluating hydrophobic contact burial, Coulombic salt-bridge pairing, and steric clash penalties across the computed contact boundary.
Can I export structural coordinates for molecular dynamics?
Yes. Clicking “Export Structure & Report” generates a comprehensive JSON bundle containing atomic coordinate backbones, per-residue pLDDT confidence records, and pairwise interface contact matrices.