A novel deep learning anomaly detection pipeline for predicting pancreatic ductal adenocarcinoma from DNA methylation data
Intelligent Medicine · 2026-10 · 6 (5) · 529-541
DOI: 10.1016/j.imed.2026.09.005 ↗Kurz gefasst
Fachartikel in der Zeitschrift Intelligent Medicine, 2026-10. Verfasst von Lois Riobó-Mayo und 10 weiteren Autor:innen. Die Arbeit nennt 60 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Carlos III Health Institute, Horizon 2020, Fundación Científica de la Asociación Española Contra el Cáncer.
Abstract (Originalsprache)
Background
Pancreatic ductal adenocarcinoma (PDAC) is highly lethal, largely because diagnosis typically occurs at advanced, incurable stages. Early detection via blood-based biomarkers could improve risk stratification. We investigated whether whole-blood DNA methylation patterns can identify individuals at elevated PDAC risk years before clinical diagnosis, using a deep learning-based anomaly detection framework applied to methylation array data from prospective cohorts.
Methods
Autoencoder models were trained on healthy control samples (N=27,185) from diverse sources available in the EWAS Data Hub and applied to Michaud’s prospective cohorts (blood collected between 1982-89), comprising 393 future PDAC cases stratified by cumulative time-to-diagnosis cutoffs (N=26 ≤ 3 y; N=57 ≤ 5 y; N=89 ≤ 7 y; N=138 ≤ 1 0 y) and 431 controls. Autoencoders were trained on windows of 10,000 consecutive CpGs with reconstruction error (REC) scores quantifying methylation anomalies. Predictive value was evaluated via stratified cross-validation using ROC AUC [95% CIs] and benchmarked against raw β-value-based classifiers. External validation was performed in the Framingham cohort (N=14 cases; N=1,331 controls, recruited between 1971-75). Functional enrichment of REC-flagged CpG regions was assessed using gene-set over-representation analysis against KEGG terms.
Results
In the prospective cohorts, REC-based classification achieved its highest macro AUC of 0.758 [0.696–0.820] in the ≤ 3-years-to-diagnosis subset, versus 0.547 [0.445-0.649] for the 𝛽 -value benchmark (ΔAUC = 0.221, p = 0.010). External validation yielded an AUC of 0.614 [0.452–0.771]. REC-flagged regions were enriched in PDAC-relevant pathways, including p53 signaling and pancreatic cancer.
Conclusion
Whole-blood DNA methylation harbours prediagnostic PDAC-associated anomalies that are detectable through deep learning-based reconstruction error modelling but not through conventional β-value analysis.
Förderung
- Carlos III Health Institute
- Horizon 2020
- Fundación Científica de la Asociación Española Contra el Cáncer
- Agence Nationale de la Recherche
- Fondazione Regionale per la Ricerca Biomedica
- Horizon 2020 Framework Programme
Zitieren
APA
Riobó-Mayo, L., Socolovsky, M., Przybyłowicz, P., Taryma-Leśniak, O., Wojdacz, T. K., Oliveri, S., Cincidda, C., Buscail, L., Balacescu, O., Obón-Santacana, M., & Moreno, V. (2026). A novel deep learning anomaly detection pipeline for predicting pancreatic ductal adenocarcinoma from DNA methylation data. Intelligent Medicine, 6(5), 529-541. https://doi.org/10.1016/j.imed.2026.09.005
Vancouver
Riobó-Mayo L, Socolovsky M, Przybyłowicz P, Taryma-Leśniak O, Wojdacz TK, Oliveri S, et al. A novel deep learning anomaly detection pipeline for predicting pancreatic ductal adenocarcinoma from DNA methylation data. Intelligent Medicine. 2026;6(5):529-541. doi:10.1016/j.imed.2026.09.005
BibTeX
@article{RiobMayo2026,
author = {Riobó-Mayo, Lois and Socolovsky, Miguel and Przybyłowicz, Patrycja and Taryma-Leśniak, Olga and Wojdacz, Tomasz K. and Oliveri, Serena and Cincidda, Clizia and Buscail, Louis and Balacescu, Ovidiu and Obón-Santacana, Mireia and Moreno, Victor},
title = {A novel deep learning anomaly detection pipeline for predicting pancreatic ductal adenocarcinoma from DNA methylation data},
journal = {Intelligent Medicine},
year = {2026},
volume = {6},
number = {5},
pages = {529--541},
doi = {10.1016/j.imed.2026.09.005}
}
EUVIMED