What question did this study set out to answer?

This research aims to evaluate the predictive power of Luoshu Convex-Hull Concavity for protein druggability.

May 9, 2026Open Access

Luoshu Convex-Hull Concavity Predicts Protein Druggability with AUC = 0.957

Key Points

This research aims to evaluate the predictive power of Luoshu Convex-Hull Concavity for protein druggability.
Evaluated 24 DUD-E benchmark proteins
Calculated Luoshu Trigram Transition Energy residues using C-alpha coordinates
Achieved results using deterministic algorithms without machine learning
CHC predicts druggability with AUC = 0.957 (p = 0.0001)
CHC alone achieves AUC = 0.871
Composite score with golden-ratio weighting reaches AUC = 0.957

Abstract

We demonstrate that Luoshu Convex-Hull Concavity (CHC) predicts protein druggability with AUC = 0. 957 (p = 0. 0001) on 24 DUD-E benchmark proteins, substantially exceeding the SiteMap Dscore benchmark of AUC ~= 0. 74 (+0. 217). CHC is the maximum interior burial depth of a convex hull constructed from the top-30% highest Luoshu Trigram Transition Energy (LTE) residues. CHC alone achieves AUC = 0. 871. Combined with LTE-top10 via golden-ratio weighting (0. 618: 0. 382), the composite score reaches AUC = 0. 957. The method requires only C-alpha coordinates and backbone dihedral angles, runs in O (n²) time (<1 second per protein), and uses no machine learning, no molecular dynamics, no ligand information, and no evolutionary data. All computations are deterministic and fully reproducible from six equations. Reference code (druggabilitydude. py) included. Extends doi: 10. 5281/zenodo. 20073303

Ask AI

Helpful

Bookmark

View Full Paper