EUVIMEDEUVIMED entdecken
Crossref

KANBind as a diagnostic probe for DNA-binding protein prediction: A prevalence-calibrated reality check under strict homology control

Qipeng Wen iD, Shaohua Jiang iD, Yiwen Zhang iD

Journal of Bioinformatics and Computational Biology · 2026-04 · 24 (02)

DOI: 10.1142/s0219720026710022 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Bioinformatics and Computational Biology, online erschienen am 21.04.2026. Verfasst von Qipeng Wen und 2 weiteren Autor:innen. Beteiligte Einrichtung: College of Information Science and Engineering, Hunan Normal University, Changsha, P. R. China. Hauptthemen laut MEDLINE: DNA-Binding Proteins, Computational Biology. Die Arbeit nennt 33 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Deep learning reports over 90% DNA-binding protein (DBP) prediction performance on common benchmarks, but these results are usually obtained on balanced test sets and may not translate to proteome-wide scans with extreme class imbalance. Here, we use KANBind as a diagnostic probe to stress-test sequence-based DBP prediction under strict homology control and realistic prevalence. Evaluated on the homology-controlled HBTD benchmark with prevalence-calibrated reporting, KANBind achieves a calibrated precision of 0.0558 at a realistic bacterial prevalence ([Formula: see text]), implying an expected false discovery rate (FDR) of 94.42%. In a proteome-scale scan, this corresponds to approximately 95 false positives per 100 predicted DBPs. Interpretability analysis indicates that predictions are driven mainly by coarse physicochemical cues such as electrostatics, which may be necessary for DNA binding but are insufficient to determine DBP function. Together, these results suggest that apparent benchmark gains can be dominated by homology leakage and evaluation on balanced sets rather than by generalizable functional rules, motivating stress-test benchmarks with strict homology control and realistic negative backgrounds.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

DNA-Binding ProteinsComputational BiologyDeep LearningPrediction Algorithms

Stichwörter der Autor:innen

DNA-binding protein predictionKolmogorov–Arnold networkclass imbalancecritical benchmarkingfalse discovery ratehomology controlprotein language model

Substanzen

DNA-Binding Proteins

Autor:innen und Institutionen

  • College of Information Science and Engineering, Hunan Normal University, Changsha, P. R. China.

ORCID-Profile: Qipeng Wen, Shaohua Jiang, Yiwen Zhang

Zitieren

APA

Wen, Q., Jiang, S., & Zhang, Y. (2026). KANBind as a diagnostic probe for DNA-binding protein prediction: A prevalence-calibrated reality check under strict homology control. Journal of Bioinformatics and Computational Biology, 24(02). https://doi.org/10.1142/s0219720026710022

Vancouver

Wen Q, Jiang S, Zhang Y. KANBind as a diagnostic probe for DNA-binding protein prediction: A prevalence-calibrated reality check under strict homology control. Journal of Bioinformatics and Computational Biology. 2026;24(02). doi:10.1142/s0219720026710022

BibTeX
@article{Wen2026,
  author = {Wen, Qipeng and Jiang, Shaohua and Zhang, Yiwen},
  title = {KANBind as a diagnostic probe for DNA-binding protein prediction: A prevalence-calibrated reality check under strict homology control},
  journal = {Journal of Bioinformatics and Computational Biology},
  year = {2026},
  volume = {24},
  number = {02},
  doi = {10.1142/s0219720026710022}
}
Kostenlos starten →