Back

SILCS-Guided Feature Encoding Expands Ligand Recognition at an HBV Core Protein Interface

Fan, Z.; Jia, R.; Lynch, D. L.; Pavlova, A.; McShan, A. C.; Gumbart, J. C.

2026-08-20 biophysics
10.64898/2026.08.16.745131 bioRxiv
Show abstract

Hepatitis B virus (HBV) infection depends on coordinated capsid assembly and virion production. A hydrophobic pocket at the intradimer interface of the HBV core protein has been linked to secretion phenotypes and shown to bind small molecules, but its interaction landscape remains poorly defined. Here, we combine atomistic molecular dynamics (MD) simulations, fragment mapping, pharmacophore modeling, and biophysical binding assays to characterize this pocket and identify ligands with binding modes that extend beyond the known pocket. This workflow narrowed an initial library of approximately 4.7 million compounds to eight candidates for experimental evaluation by saturation transfer difference (STD) NMR and surface plasmon resonance (SPR). Of these eight, compound B3 showed detectable STD NMR signals, concentration-dependent SPR binding, and an MD-supported binding mode that retained hydrophobic-pocket anchoring while also extending toward the spike-proximal loop. Together, these results illustrate how dynamic fragment mapping can identify ligand candidates that engage broader interaction landscapes at protein-interface pockets.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.