EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

CrossAttOmics: multiomics data integration with cross-attention

Aurélien Beaude iD, Franck Augé iD, Farida Zehraoui iD, Blaise Hanczar iD

Bioinformatics · 02.06.2025 · 41 (6)

DOI: 10.1093/bioinformatics/btaf302 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Bioinformatics, online erschienen am 13.05.2025. Verfasst von Aurélien Beaude und 3 weiteren Autor:innen. Beteiligte Einrichtung: Université Paris-Saclay, Univ Evry, IBISC, Evry-Courcouronnes 91020, France. Hauptthemen laut MEDLINE: Genomics, Computational Biology, Deep Learning. Die Arbeit wurde bisher 20-mal zitiert (Crossref) und nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: CIFRE.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Motivation Advances in high throughput technologies enabled large access to various types of omics. Each omics provides a partial view of the underlying biological process. Integrating multiple omics layers would help have a more accurate diagnosis. However, the complexity of omics data requires approaches that can capture complex relationships. One way to accomplish this is by exploiting the known regulatory links between the different omics, which could help in constructing a better multimodal representation. Results In this article, we propose CrossAttOmics, a new deep-learning architecture based on the cross-attention mechanism for multiomics integration. Each modality is projected in a lower dimensional space with its specific encoder. Interactions between modalities with known regulatory links are computed in the feature representation space with cross-attention. The results of different experiments carried out in this article show that our model can accurately predict the types of cancer by exploiting the interactions between multiple modalities. CrossAttOmics outperforms other methods when there are few paired training examples. Our approach can be combined with attribution methods like LRP to identify which interactions are the most important. Availability and implementation The code is available at https://github.com/Sanofi-Public/CrossAttOmics and https://doi.org/10.5281/zenodo.15065928. TCGA data can be downloaded from the Genomic Data Commons Data Portal. CCLE data can be downloaded from the depmap portal.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

GenomicsComputational BiologyDeep LearningHumansNeoplasmsSoftwareAlgorithmsMultiomics

Autor:innen und Institutionen

  • Université Paris-Saclay, Univ Evry, IBISC, Evry-Courcouronnes 91020, France.
  • Sanofi R&D, Translational Precision Medicine, Vitry-sur-Seine 94400, France.

ORCID-Profile: Aurélien Beaude, Franck Augé, Farida Zehraoui, Blaise Hanczar

Förderung

  • CIFRE

Zitieren

APA

Beaude, A., Augé, F., Zehraoui, F., & Hanczar, B. (2025). CrossAttOmics: multiomics data integration with cross-attention. Bioinformatics, 41(6). https://doi.org/10.1093/bioinformatics/btaf302

Vancouver

Beaude A, Augé F, Zehraoui F, Hanczar B. CrossAttOmics: multiomics data integration with cross-attention. Bioinformatics. 2025;41(6). doi:10.1093/bioinformatics/btaf302

BibTeX
@article{Beaude2025,
  author = {Beaude, Aurélien and Augé, Franck and Zehraoui, Farida and Hanczar, Blaise},
  title = {CrossAttOmics: multiomics data integration with cross-attention},
  journal = {Bioinformatics},
  year = {2025},
  volume = {41},
  number = {6},
  doi = {10.1093/bioinformatics/btaf302}
}
Kostenlos starten →