ERFMTDA: Predicting tsRNA–disease associations using an enhanced rotative factorization machine

tRNA-derived small RNAs (tsRNAs) have emerged as a novel class of regulatory molecules implicated in the pathogenesis of numerous human diseases, positioning them as promising biomarkers and therapeutic targets. Existing computational methods provide a cost-effective alternative to experimental method, but they tend to ignore biological attributes and complex feature interactions. To overcome these limitations, we propose ERFMTDA, an enhanced rotative factorization machine framework for predicting potential tsRNA-disease associations. ERFMTDA explicitly models complex interactions among heterogeneous biological features while integrating latent structural representations derived from the global association matrix. In addition, a biologically informed negative sampling strategy based on motif-level sequence similarity is introduced to improve the reliability of negative samples. Extensive experiments demonstrate that ERFMTDA consistently surpasses the other eleven state-of-the-art methods. Two case studies on diabetic retinopathy and hepatocellular carcinoma further corroborate the model’s ability to prioritize biologically meaningful tsRNA–disease associations.Author summary: tRNA-derived small RNAs (tsRNAs) are a class of regulatory molecules that play important roles in diverse biological processes and human diseases. Identifying disease-associated tsRNAs is crucial for understanding disease mechanisms and developing potential diagnostic and therapeutic strategies. As experimental validation of tsRNA–disease associations is costly and time-consuming, computational methods have become essential for tsRNA–disease association prediction. However, most existing methods focus on network topology and fail to incorporate explicit biological features and complex feature interactions. To address these limitations, we propose ERFMTDA, a computational framework that integrates biological features with global structural information. The framework further incorporates a motif-based negative sampling strategy to reduce false negative samples. By capturing high-order and non-linear interactions among heterogeneous features, ERFMTDA effectively models the complex relationships between tsRNAs and diseases. Extensive validation experiments and two case studies on diabetic retinopathy and hepatocellular carcinoma demonstrate that ERFMTDA can reliably prioritize biologically relevant tsRNA–disease associations, providing a practical tool for guiding future experimental studies.