Interpreting convolutional neural networks in population genetics
Molecular Biology and Evolution · 03.10.2026 · 43 (10)
DOI: 10.1093/molbev/msag256 ↗Kurz gefasst
Fachartikel in der Zeitschrift Molecular Biology and Evolution, online erschienen am 05.10.2026. Verfasst von Huiting Xu und 6 weiteren Autor:innen. Beteiligte Einrichtung: Institute for Bioinformatics and Medical Informatics (IBMI), University of Tübingen, 72076 Tübingen, Germany. Hauptthemen laut MEDLINE: Convolutional Neural Networks, Genetics, Population. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 94 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Reinhard Frank-Stiftung, Deutsche Forschungsgemeinschaft, NHGRI NIH HHS.
Abstract (Originalsprache)
Machine learning approaches have become a powerful alternative to traditional methods in population genetics. Convolutional neural networks (CNNs) in particular have been successful in inferring natural selection, recombination rates, introgression, dispersal distances, and effective population size changes. One limitation of CNNs and other deep learning methods is that they can be difficult to interpret. When they have been shown to be as or more successful than summary-statistic-based methods, what are they learning? Here, we investigate CNNs from two different methods: the pg-gan discriminator for identifying real vs. simulated data, and two networks trained to detect selective sweeps. We first compute correlations between learned network features and traditional summary statistics, then assess whether summary statistics can be predicted from the learned features. To understand the learned features, we compute feature importance through SHAP values and feature groupings through dimensionality reduction. Finally, we use decision trees and random forests to build an interpretable “model-of-the-model”. Our results reveal that some CNN architectures can implicitly compute summary statistics such as pairwise heterozygosity, while statistics such as the site frequency spectrum are less similar to the network’s learned features. We find that long-range linkage disequilibrium is readily approximated by the networks and may be more efficiently computed by CNNs than traditional methods (which are quadratic in the number of sites). Overall, this work contributes to the interpretability of deep learning methods in population genetics by clarifying the relationships between model architecture, learned network features, established summary statistics, and predicted evolutionary parameters.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Institute for Bioinformatics and Medical Informatics (IBMI), University of Tübingen, 72076 Tübingen, Germany.
- Department of Biology, University of Pennsylvania, Philadelphia, PA, USA.
- Department of Computer Science, Haverford College, Haverford, PA, USA.
- Department of Genetics, University of North Carolina at Chapel Hill, USA.
- Cluster of Excellence "Controlling Microbes to Fight Infections", Mathematical and Computational Population Genetics, University of Tübingen, 72076 Tübingen, Germany.
- Cluster of Excellence "Machine Learning: New Perspectives for Science", University of Tübingen, Tübingen, Germany.
ORCID-Profile: Leon Zong, Daniel R Schrider, Franz Baumdicker, Sara Mathieson
Förderung
- Reinhard Frank-Stiftung
- Deutsche Forschungsgemeinschaft
- NHGRI NIH HHS
- National Institutes of Health (NIH)
- NIGMS NIH HHS
- NIH
Zitieren
APA
Xu, H., Zong, L., Ray, D. D., Lei, L., Schrider, D. R., Baumdicker, F., & Mathieson, S. (2026). Interpreting convolutional neural networks in population genetics. Molecular Biology and Evolution, 43(10). https://doi.org/10.1093/molbev/msag256
Vancouver
Xu H, Zong L, Ray DD, Lei L, Schrider DR, Baumdicker F, et al. Interpreting convolutional neural networks in population genetics. Molecular Biology and Evolution. 2026;43(10). doi:10.1093/molbev/msag256
BibTeX
@article{Xu2026,
author = {Xu, Huiting and Zong, Leon and Ray, Dylan D and Lei, Lei and Schrider, Daniel R and Baumdicker, Franz and Mathieson, Sara},
title = {Interpreting convolutional neural networks in population genetics},
journal = {Molecular Biology and Evolution},
year = {2026},
volume = {43},
number = {10},
doi = {10.1093/molbev/msag256}
}
EUVIMED