Prediction of circRNA-miRNA interaction sites based on fully convolutional neural network and gated recurrent unit
*Correspondence to:
Zhen Shen, School of Computer and Software, Nanyang Institute of Technology, Nanyang 473004, Henan, China.
E-mail: zzuliszhen@163.com
Comput Biomed. 2026;1:202620. 10.70401/cbm.2026.0025
Received: June 18, 2026Accepted: September 20, 2026Published: September 20, 2026
This manuscript is made available in its unedited form to allow early access to the
reported findings. Further editing will be completed before final publication. As
such,
the content may include errors, and standard legal disclaimers are applicable.
Abstract
Aims: Circular RNAs (circRNAs) act as vital miRNA sponges and participate in the pathogenesis of various complex diseases. Most existing computational methods focus on sequence-level prediction of circRNA-miRNA interactions, while few tools can accurately locate interaction sites at the nucleotide level.
Methods: In this study, we propose CMIFCG, a novel model combining fully convolutional neural networks (FCN) and bidirectional gated recurrent units (BiGRU) for nucleotide-level prediction of circRNA-miRNA interaction sites. We adopted one-hot encoding for sequence preprocessing, utilized stacked convolutional modules to extract high-order features, and integrated BiGRU to capture sequential association features. The decoding module with skip connections restores feature maps to the original sequence size for precise site prediction. We constructed four datasets with different sequence lengths and negative samples to evaluate model performance.
Results: Experimental results show that CMIFCG achieves optimal performance than other baseline models, and excessive module stacking causes performance degradation. Motif analysis confirms that the predicted motifs match well with known miRNA sequences. Partial predicted interactions are further verified via public databases.
Conclusion: In summary, CMIFCG model enables effective fine-grained identification of circRNA-miRNA interaction sites, providing a new tool for exploring circRNA regulatory mechanisms and disease-related molecular pathways.
Keywords
circRNA-miRNA interaction, fully convolutional neural network, gated recurrent unit, motif analysis
References
-
3. Niu M, Chen Y, Wang C, Zou Q, Xu L. Computational approaches for circRNA-disease association prediction: A review. Front Comput Sci. 2024;19(4):194904.[DOI]
-
4. Zhang J, Zhao F. Circular RNA discovery with emerging sequencing and deep learning technologies. Nat Genet. 2025;57(5):1089-1102.[DOI]
-
6. Panda AC. Circular RNAs act as miRNA sponges. In: Xiao J, editor. Circular RNAs: Biogenesis and functions. Singapore: Springer; 2018. p. 67-79.[DOI]
-
7. Li S, Cui Z, Gao M, Shan Y, Ren Y, Zhao Y, et al. Hsa_circ_0072088 promotes non-small cell lung cancer progression through modulating miR-1270/TOP2A axis. Cancer Cell Int. 2025;25(1):156.[DOI]
-
9. Yao H, Song M, Zhang H, Li Y, Chen D, Li Y, et al. NSUN2/ALYREF-medicated M5C-modified circRNA505 regulates the proliferation, differentiation, and glycolysis of antler chondrocytes via the miRNA-127/p53 axis and LDHA. Int J Biol Macromol. 2025;309:142527.[DOI]
-
10. Schmidt D, Wilson MD, Spyrou C, Brown GD, Hadfield J, Odom DT. ChIP-seq: Using high-throughput sequencing to discover protein–DNA interactions. Methods. 2009;48(3):240-248.[DOI]
-
11. Adams G. A beginner’s guide to RT-PCR, qPCR and RT-qPCR. Biochemist. 2020;42(3):48-53.[DOI]
-
12. Wei MM, Wang L, Zhao BW, Su XR, You ZH, Huang DS. Integrating transformer and graph attention network for circRNA-miRNA interaction prediction. IEEE J Biomed Health Inform. 2025;29(8):6105-6113.[DOI]
-
14. Cao C, Li M, Wang C, Xu L, Zou Q, Wang Y, et al. DGCLCMI: A deep graph collaboration learning method to predict circRNA-miRNA interactions. BMC Biol. 2025;23(1):104.[DOI]
-
15. Li YC, You ZH, Yu CQ, Wang L, Hu L, Hu PW, et al. DeepCMI: A graph-based model for accurate prediction of circRNA–miRNA interactions with multiple information. Brief Funct Genomics. 2024;23(3):276-285.[DOI]
-
16. Wei M, Wang L, Su X, Zhao B, You Z. Multi-hop graph structural modeling for cancer-related circRNA-miRNA interaction prediction. Pattern Recognit. 2026;170:112078.[DOI]
-
17. Liu L, Wei Y, Zhang Q, Zhao Q. SSCRB: Predicting circRNA-RBP interaction sites using a sequence and structural feature-based attention model. IEEE J Biomed Health Inform. 2024;28(3):1762-1772.[DOI]
-
20. Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation. In: 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2015 Jun 7-12; Boston, USA. Piscataway: IEEE; 2015. p. 3431-3440.[DOI]
-
25. Liu N, Zhang Y. A deep learning approach based on molecular graph features and residual blocks to predict interaction sites between CircRNA and RBP. Biochem Biophys Res Commun. 2025;785:152655.[DOI]
-
26. Yang Y, Hou Z, Ma Z, Li X, Wong KC. iCircRBP-DHN: Identification of circRNA-RBP interaction sites using deep hierarchical network. Brief Bioinform. 2021;22(4):bbaa274.[DOI]
-
29. Zhang Y, Liu S, Wu C, Gao X, Zhao H, Li O, et al. Exosomal circ_0001583 drives glioblastoma cell advancement through the miR-647/CKAP2L pathway. Mol Neurobiol. 2025;62(8):10687-10706.[DOI]
-
30. CHEN B, FENG H, ZOU C-G et al. Effects of circTRIM33-12 on proliferation, apoptosis and epithelial-mesenchymal transition of brain glioma cells by regulating miR-191/DAB2 axis. Journal of Regional Anatomy and Operative Surgery. 2024: 36-43. DOI: 10.11659/jjssx.11611E022142
-
36. Aslani S, Kalantary-Charvadeh A, Morovat P, Abbasalipourkabir R, Nourmohammadi I, Emami Razavi A, et al. Integrated bioinformatics and tissue-based validation reveal the oncogenic role of hsa_circ_0043256 and hsa_circ_0004789 in gastric cancer. Biochem Biophys Rep. 2026;45:102442.
-
42. Yin W, Zhang Z, Zhang S, He L, Zhang R, Jiang R, et al. ERNIE-RNA: An RNA language model with structure-enhanced representations. Nat Commun. 2025;16:10076.[DOI]
Copyright
© The Author(s) 2026. This is an Open Access article licensed under a Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, sharing, adaptation, distribution and reproduction in any medium or format, for any purpose, even commercially, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
Publisher’s Note
Science Exploration remains a neutral stance on jurisdictional claims in published
maps
and institutional affiliations. The views expressed in this article are solely those
of
the author(s) and do not reflect the opinions of the Editors or the publisher.
Share And Cite
Science Exploration Style
Qiu Y, Li Y, Zhao Y, Zhou P, Shen Z. Prediction of circRNA-miRNA interaction sites based on fully convolutional neural network and gated recurrent unit. Comput Biomed. 2026;1:202620. https://doi.org/10.70401/cbm.2026.0025
Tips
Copy completed.
Submit a Manuscript
Author Instructions
Cite this Article
Article Metrics
0
View
0
Download
Cited
Article Updates
Get citation
Science Exploration Style
Qiu Y, Li Y, Zhao Y, Zhou P, Shen Z. Prediction of circRNA-miRNA interaction sites based on fully convolutional neural network and gated recurrent unit. Comput Biomed. 2026;1:202620. https://doi.org/10.70401/cbm.2026.0025
copy
Share Link
copy