LncMamba: A deep learning model for LncRNA localization prediction based on the Mamba model

Accurate prediction of long non-coding RNA (LncRNA) subcellular localization is crucial for understanding its biological functions. In this study, we proposed a novel deep learning framework, LncMamba, which utilizes a two-layer FPN network for multi-scale feature extraction and introduces the Mamba network to LncRNA localization prediction tasks for the first time. Based on this, we improved the localization-specific attention mechanism, allowing the model to focus more effectively on key sequence motifs related to localization. Additionally, through statistical analysis of localization motifs, we revealed the potential relationship between nucleotide motifs and LncRNA subcellular localization. The code is available at: https://anonymous.4open.science/r/LncMamba-731F.