Detecting genetic interactions with visible neural networks
Abstract Non-linear interactions among single nucleotide polymorphisms (SNPs), genes, and pathways play an important role in human diseases, but identifying these interactions is a challenging task. Neural networks are state-of-the-art predictors in many domains due to their ability to analyze big d...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-06-01
|
| Series: | Communications Biology |
| Online Access: | https://doi.org/10.1038/s42003-025-08157-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849434159929360384 |
|---|---|
| author | Arno van Hilten Federico Melograna Bowen Fan Wiro Niessen Kristel van Steen Gennady Roshchupkin |
| author_facet | Arno van Hilten Federico Melograna Bowen Fan Wiro Niessen Kristel van Steen Gennady Roshchupkin |
| author_sort | Arno van Hilten |
| collection | DOAJ |
| description | Abstract Non-linear interactions among single nucleotide polymorphisms (SNPs), genes, and pathways play an important role in human diseases, but identifying these interactions is a challenging task. Neural networks are state-of-the-art predictors in many domains due to their ability to analyze big data and model complex patterns, including non-linear interactions. In genetics, visible neural networks are popular as they provide insight into the most important SNPs, genes, and pathways for prediction. Visible neural networks use prior knowledge (e.g., gene and pathway annotations) to define node connections in the network, making them sparse and interpretable. Currently, most of these networks provide measures for the importance of SNPs, genes, and pathways but do not provide information about interactions. In this paper, we explore different methods to detect non-linear interactions with visible neural networks. We adapt and speed up existing methods, create a comprehensive benchmark with simulated data from GAMETES and EpiGEN, and demonstrate that these methods can extract multiple types of interactions from trained neural networks. Finally, we apply these methods to a genome-wide case-control study of inflammatory bowel disease and find high consistency of the epistasis pairs candidates between interpretation methods. The follow-up association test on these candidates identifies seven significant epistasis pairs. |
| format | Article |
| id | doaj-art-ccce4b3c8e5b426bb2e753031e530ef9 |
| institution | Kabale University |
| issn | 2399-3642 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Communications Biology |
| spelling | doaj-art-ccce4b3c8e5b426bb2e753031e530ef92025-08-20T03:26:47ZengNature PortfolioCommunications Biology2399-36422025-06-018111210.1038/s42003-025-08157-xDetecting genetic interactions with visible neural networksArno van Hilten0Federico Melograna1Bowen Fan2Wiro Niessen3Kristel van Steen4Gennady Roshchupkin5Department of Radiology and Nuclear Medicine, Erasmus MCBIO3 - Laboratory for Systems Medicine, Department of Human Genetics, KU LeuvenDepartment of Biosystems Science and Engineering, ETH ZurichDepartment of Radiology and Nuclear Medicine, Erasmus MCBIO3 - Laboratory for Systems Medicine, Department of Human Genetics, KU LeuvenDepartment of Radiology and Nuclear Medicine, Erasmus MCAbstract Non-linear interactions among single nucleotide polymorphisms (SNPs), genes, and pathways play an important role in human diseases, but identifying these interactions is a challenging task. Neural networks are state-of-the-art predictors in many domains due to their ability to analyze big data and model complex patterns, including non-linear interactions. In genetics, visible neural networks are popular as they provide insight into the most important SNPs, genes, and pathways for prediction. Visible neural networks use prior knowledge (e.g., gene and pathway annotations) to define node connections in the network, making them sparse and interpretable. Currently, most of these networks provide measures for the importance of SNPs, genes, and pathways but do not provide information about interactions. In this paper, we explore different methods to detect non-linear interactions with visible neural networks. We adapt and speed up existing methods, create a comprehensive benchmark with simulated data from GAMETES and EpiGEN, and demonstrate that these methods can extract multiple types of interactions from trained neural networks. Finally, we apply these methods to a genome-wide case-control study of inflammatory bowel disease and find high consistency of the epistasis pairs candidates between interpretation methods. The follow-up association test on these candidates identifies seven significant epistasis pairs.https://doi.org/10.1038/s42003-025-08157-x |
| spellingShingle | Arno van Hilten Federico Melograna Bowen Fan Wiro Niessen Kristel van Steen Gennady Roshchupkin Detecting genetic interactions with visible neural networks Communications Biology |
| title | Detecting genetic interactions with visible neural networks |
| title_full | Detecting genetic interactions with visible neural networks |
| title_fullStr | Detecting genetic interactions with visible neural networks |
| title_full_unstemmed | Detecting genetic interactions with visible neural networks |
| title_short | Detecting genetic interactions with visible neural networks |
| title_sort | detecting genetic interactions with visible neural networks |
| url | https://doi.org/10.1038/s42003-025-08157-x |
| work_keys_str_mv | AT arnovanhilten detectinggeneticinteractionswithvisibleneuralnetworks AT federicomelograna detectinggeneticinteractionswithvisibleneuralnetworks AT bowenfan detectinggeneticinteractionswithvisibleneuralnetworks AT wironiessen detectinggeneticinteractionswithvisibleneuralnetworks AT kristelvansteen detectinggeneticinteractionswithvisibleneuralnetworks AT gennadyroshchupkin detectinggeneticinteractionswithvisibleneuralnetworks |