![[2602.22236] CrossLLM-Mamba: Multimodal State Space Fusion of LLMs for RNA Interaction Prediction](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.22236] CrossLLM-Mamba: Multimodal State Space Fusion of LLMs for RNA Interaction Prediction
Summary
The article presents CrossLLM-Mamba, a novel framework for RNA interaction prediction that utilizes multimodal state space fusion of large language models, achieving state-of-the-art performance in various biological interaction categories.
Why It Matters
This research is significant as it addresses the limitations of existing RNA interaction prediction methods by introducing a dynamic modeling approach. This advancement could enhance our understanding of cellular regulation and improve drug discovery processes, making it relevant for both academic research and pharmaceutical applications.
Key Takeaways
- CrossLLM-Mamba reformulates RNA interaction prediction as a state-space alignment problem.
- The framework allows for dynamic modeling of interactions through hidden state propagation.
- Achieves state-of-the-art performance on the RPI1460 benchmark with an MCC of 0.892.
- Incorporates Gaussian noise injection and Focal Loss for improved robustness.
- Demonstrates the potential of state-space modeling in multimodal biological predictions.
Quantitative Biology > Genomics arXiv:2602.22236 (q-bio) [Submitted on 23 Feb 2026] Title:CrossLLM-Mamba: Multimodal State Space Fusion of LLMs for RNA Interaction Prediction Authors:Rabeya Tus Sadia, Qiang Ye, Qiang Cheng View a PDF of the paper titled CrossLLM-Mamba: Multimodal State Space Fusion of LLMs for RNA Interaction Prediction, by Rabeya Tus Sadia and 2 other authors View PDF HTML (experimental) Abstract:Accurate prediction of RNA-associated interactions is essential for understanding cellular regulation and advancing drug discovery. While Biological Large Language Models (BioLLMs) such as ESM-2 and RiNALMo provide powerful sequence representations, existing methods rely on static fusion strategies that fail to capture the dynamic, context-dependent nature of molecular binding. We introduce CrossLLM-Mamba, a novel framework that reformulates interaction prediction as a state-space alignment problem. By leveraging bidirectional Mamba encoders, our approach enables deep ``crosstalk'' between modality-specific embeddings through hidden state propagation, modeling interactions as dynamic sequence transitions rather than static feature overlaps. The framework maintains linear computational complexity, making it scalable to high-dimensional BioLLM embeddings. We further incorporate Gaussian noise injection and Focal Loss to enhance robustness against hard-negative samples. Comprehensive experiments across three interaction categories, RNA-protein, RNA-small molecule, a...