EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Bias-aware training and evaluation of link prediction algorithms in network biology

Serhan Yılmaz iD, Kaan Yorgancioglu, Mehmet Koyutürk

Proceedings of the National Academy of Sciences · 17.06.2025 · 122 (24)

DOI: 10.1073/pnas.2416646122 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the National Academy of Sciences, online erschienen am 10.06.2025. Verfasst von Serhan Yılmaz und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer and Data Sciences, Case Western Reserve University, Cleveland, OH 44106. Hauptthemen laut MEDLINE: Algorithms, Computational Biology. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 61 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: NLM NIH HHS, HHS | NIH | U.S. National Library of Medicine (NLM).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

For biomedical applications, new link prediction algorithms are continuously being developed. These algorithms are typically evaluated computationally, using test sets generated by sampling the edges uniformly at random. However, as we demonstrate, this evaluation approach introduces a bias toward “rich nodes,” i.e., those with higher degrees in the network. More concerningly, this bias persists even when different network snapshots are used for evaluation, as recommended in the machine learning community. This creates a cycle in research where newly developed algorithms generate more knowledge on well-studied biological entities while understudied entities are commonly overlooked. To overcome this issue, we propose a weighted validation setting specifically focusing on low-degree nodes and present AWARE strategies to facilitate bias-aware training and evaluation of link prediction algorithms. These strategies can help researchers gain better insights from computational evaluations and promote the development of new algorithms focusing on novel findings and understudied proteins.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

AlgorithmsComputational BiologyMachine LearningHumansPrediction Algorithms

Stichwörter der Autor:innen

biasgraph machine learningnetwork biologyprotein–protein interactionvalidation

Autor:innen und Institutionen

  • Department of Computer and Data Sciences, Case Western Reserve University, Cleveland, OH 44106.

ORCID-Profile: Serhan Yılmaz

Förderung

  • NLM NIH HHS
  • HHS | NIH | U.S. National Library of Medicine (NLM)

Zitieren

APA

Yılmaz, S., Yorgancioglu, K., & Koyutürk, M. (2025). Bias-aware training and evaluation of link prediction algorithms in network biology. Proceedings of the National Academy of Sciences, 122(24). https://doi.org/10.1073/pnas.2416646122

Vancouver

Yılmaz S, Yorgancioglu K, Koyutürk M. Bias-aware training and evaluation of link prediction algorithms in network biology. Proceedings of the National Academy of Sciences. 2025;122(24). doi:10.1073/pnas.2416646122

BibTeX
@article{Ylmaz2025,
  author = {Yılmaz, Serhan and Yorgancioglu, Kaan and Koyutürk, Mehmet},
  title = {Bias-aware training and evaluation of link prediction algorithms in network biology},
  journal = {Proceedings of the National Academy of Sciences},
  year = {2025},
  volume = {122},
  number = {24},
  doi = {10.1073/pnas.2416646122}
}
Kostenlos starten →