EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

BioLM-Score: Language-Prior Conditioned Probabilistic Geometric Potentials for Protein-Ligand Scoring

Zhangfan Yang, Baoyun Chen, Dong Xu, Jia Wang, Ruibin Bai, Zexuan Zhu, Junkai Ji

Bioinformatics · 14.11.2026

DOI: 10.1093/bioinformatics/btag839 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Bioinformatics, online erschienen am 14.11.2026. Verfasst von Zhangfan Yang und 6 weiteren Autor:innen. Beteiligte Einrichtung: School of Artificial Intelligence, Shenzhen University, Shenzhen, China. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Motivation Protein-ligand scoring must address affinity prediction, pose selection, and virtual screening, but these objectives are not necessarily optimized together. Geometric likelihood models may generalize poorly when training-set structural coverage is limited. We test whether pretrained biomolecular representations improve scoring under a controlled geometric baseline. Results BioLM-Score integrates frozen protein and molecular language-model embeddings into a residue-atom distance-likelihood framework via gated graph convolutions. We evaluate geometry-only, jointly trained, and two-stage fine-tuned variants, using the jointly trained variant for the primary controlled comparisons. All three DEKOIS 2.0 screening endpoints improve over the matched geometric baseline with target-level paired support, but affinity scoring and ranking do not improve uniformly. Under a time-split protocol, scoring is significantly lower and ranking is also lower than the geometric baseline, while screening gains persist. On BDB2020+, the difference in affinity correlation between BioLM-Score and GenScore narrows after LP-PDBBind-compatible retraining. Shuffling protein embeddings at inference reduces the screening advantage. However, retrained shuffle controls recover much of the performance, so correct alignment is not established as necessary. BioLM-Score strengthens geometric scoring on screening tasks without retraining the language models. Availability and implementation Code, dataset-split identifiers, preprocessing and evaluation scripts, and language-model embedding-generation utilities are available at https://github.com/SZU-ADDG/BioLM-Score. An archival code snapshot and pretrained model weights are available at doi:10.5281/zenodo.21878818. Supplementary information Supplementary data are available at Bioinformatics online.

Schlagwörter

Stichwörter der Autor:innen

biomolecular language modelsdrug discoverygeometric deep learningprotein-ligand scoring

Autor:innen und Institutionen

  • School of Artificial Intelligence, Shenzhen University, Shenzhen, China.
  • School of Computer Science, University of Nottingham Ningbo China, Ningbo, China.
  • College of Computer Science and Software Engineering, Shenzhen University.
  • EasternDawn, Shenzhen, China.
  • Intelligent Computing Center, Shenzhen University, 518060 Shenzhen, China.

Zitieren

APA

Yang, Z., Chen, B., Xu, D., Wang, J., Bai, R., Zhu, Z., & Ji, J. (2026). BioLM-Score: Language-Prior Conditioned Probabilistic Geometric Potentials for Protein-Ligand Scoring. Bioinformatics. https://doi.org/10.1093/bioinformatics/btag839

Vancouver

Yang Z, Chen B, Xu D, Wang J, Bai R, Zhu Z, et al. BioLM-Score: Language-Prior Conditioned Probabilistic Geometric Potentials for Protein-Ligand Scoring. Bioinformatics. 2026. doi:10.1093/bioinformatics/btag839

BibTeX
@article{Yang2026,
  author = {Yang, Zhangfan and Chen, Baoyun and Xu, Dong and Wang, Jia and Bai, Ruibin and Zhu, Zexuan and Ji, Junkai},
  title = {BioLM-Score: Language-Prior Conditioned Probabilistic Geometric Potentials for Protein-Ligand Scoring},
  journal = {Bioinformatics},
  year = {2026},
  doi = {10.1093/bioinformatics/btag839}
}
Kostenlos starten →