SMAP: Similarity-based matrix factorization framework for inferring miRNA-disease association

Based on increasing evidence, microRNAs (miRNAs) play significant roles in various complex human diseases. Therefore, identifying the disease-related miRNAs could help to provide a stepping stone to help understand the disease pathogenesis mechanism. However, owing to the cost and complexity of clinical methods, the development of novel computational frameworks that identify miRNA-disease associations could be a great alternative. Herein, we sought to present a simple yet effective computational framework (SMAP) for identifying miRNA-disease associations by applying the recommended algorithm with miRNA and disease similarity constraints. The comprehensive and accurate similarity values were measured based on miRNA functional similarity, disease semantic similarity, and Gaussian interaction profile kernel similarity. SMAP not only exploited known miRNA-disease associations for the matrix factorization model, but also utilized the integrated similarities for miRNAs and diseases. As a result, SMAP had AUCs of 0.9227 and 0.8952 in the frameworks of global and local leave-one-out cross validation, respectively. In addition, independent case studies on two major human cancers clearly verified the comparable performance of SMAP. Overall, SMAP could serve as an effective guide to elucidating pathogenesis and etiology of human diseases and deciphering the potential disease biomarkers.