EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

When complexity does not pay: benchmarking deep learning and ensemble methods for biomarker discovery

Cyrille Mesue Njume iD, Irene Petracci, Sonia Bellini iD, Katarzyna Goljanek-Whysall, Leo R Quinlan, Agnieszka Fiszer, Barbara Borroni, Roberta Ghidoni, Asli Kumbasar, Ali Cakmak iD

Briefings in Bioinformatics · 03.05.2026 · 27 (3)

DOI: 10.1093/bib/bbag211 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Briefings in Bioinformatics, online erschienen am 08.05.2026. Verfasst von Cyrille Mesue Njume und 9 weiteren Autor:innen. Beteiligte Einrichtung: Department of Molecular Biology and Genetics, Ayazaga Campus, Istanbul Technical University, Reşitpaşa, Sarıyer, 34467 Istanbul, Turkey. Hauptthemen laut MEDLINE: Biomarkers, Deep Learning, Computational Biology. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Scientific and Technological Research Council of Turkey, EU Joint Programme-Neurodegenerative Disease Research, Scientific Research Projects Unit of Istanbul Technical University.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The integration of multi-omics data holds great promise for identifying robust and clinically relevant biomarkers, yet the increasing complexity of computational methods raises questions about their practical utility. In this study, we present a comprehensive benchmarking framework that evaluates 27 feature selection strategies and 11 predictive models across three real-world disease cohorts: Alzheimer’s disease, progressive supranuclear palsy, and breast cancer. We compare traditional machine learning, ensemble-based methods, and state-of-the-art deep learning models in terms of predictive performance, stability, and biological interpretability. Our results reveal that ensemble feature selection consistently improves robustness and accuracy, particularly for compact biomarker panels. Surprisingly, deep learning models did not outperform simpler classifiers such as logistic regression (L.Regression), support vector machines, or multilayer perceptrons, which often achieved comparable or superior results with lower computational cost and greater interpretability. Triple-omics yielded the highest validation, followed by dual-omics and then single-omics (Triple > Dual > Single). Biological validation against five independent databases confirmed the clinical relevance of the identified biomarkers, including both well-established and novel candidates. To support reproducibility and community adoption, we provide a web-based tool for applying our benchmarking pipeline. Our findings advocate for a pragmatic approach to biomarker discovery—prioritizing methodological transparency, reproducibility, and biological insight over algorithmic complexity.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

BiomarkersDeep LearningComputational BiologyHumansBenchmarkingMultiomicsEnsemble LearningAlzheimer DiseaseBreast NeoplasmsReproducibility of Results

Stichwörter der Autor:innen

biomarker discoveryensemble rank aggregationfeature selection benchmarkingintegrative bioinformaticsmulti-omics integration

Substanzen

Biomarkers

Autor:innen und Institutionen

  • Department of Molecular Biology and Genetics, Ayazaga Campus, Istanbul Technical University, Reşitpaşa, Sarıyer, 34467 Istanbul, Turkey.
  • Molecular Markers Laboratory, IRCCS Istituto Centro San Giovanni di Dio Fatebenefratelli, 25125 Brescia, Italy.
  • Discipline of Physiology, School of Medicine, University of Galway, H91 TH33 Galway, Ireland.
  • Department of Medical Biotechnology, Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14, 61-704 Poznan, Poland.
  • Department of Computer Engineering, Ayazaga Campus, Istanbul Technical University, Reşitpaşa, Sarıyer, 34467 Istanbul, Turkey.

ORCID-Profile: Cyrille Mesue Njume, Sonia Bellini, Ali Cakmak

Förderung

  • Scientific and Technological Research Council of Turkey
  • EU Joint Programme-Neurodegenerative Disease Research
  • Scientific Research Projects Unit of Istanbul Technical University
  • National Center for High-Performance Computing
  • Italian Ministry of Health-EU Joint Programme-Neurodegenerative Disease Research
  • Polish National Science Centre
  • Health Research Board
  • EU Joint Programme—Neurodegenerative Disease Research

Zitieren

APA

Njume, C. M., Petracci, I., Bellini, S., Goljanek-Whysall, K., Quinlan, L. R., Fiszer, A., Borroni, B., Ghidoni, R., Kumbasar, A., & Cakmak, A. (2026). When complexity does not pay: benchmarking deep learning and ensemble methods for biomarker discovery. Briefings in Bioinformatics, 27(3). https://doi.org/10.1093/bib/bbag211

Vancouver

Njume CM, Petracci I, Bellini S, Goljanek-Whysall K, Quinlan LR, Fiszer A, et al. When complexity does not pay: benchmarking deep learning and ensemble methods for biomarker discovery. Briefings in Bioinformatics. 2026;27(3). doi:10.1093/bib/bbag211

BibTeX
@article{Njume2026,
  author = {Njume, Cyrille Mesue and Petracci, Irene and Bellini, Sonia and Goljanek-Whysall, Katarzyna and Quinlan, Leo R and Fiszer, Agnieszka and Borroni, Barbara and Ghidoni, Roberta and Kumbasar, Asli and Cakmak, Ali},
  title = {When complexity does not pay: benchmarking deep learning and ensemble methods for biomarker discovery},
  journal = {Briefings in Bioinformatics},
  year = {2026},
  volume = {27},
  number = {3},
  doi = {10.1093/bib/bbag211}
}
Kostenlos starten →