The training dataset and the leave-one-out validation results on Baskerville et al.' data [14].

*Once again, kicking one sample out as the testing sample, the rest 28 samples are the training dataset.The four features (columns ��1��, ��2��, ��3��, and ��4��) of each miRNA are calculated based on the genomic coordinates of the miRNA, the miRNA hosting intron, and the host gene.ER represents the experimental results and PR represents the prediction results. The symbol ��+�� means high co-expression and the symbol ��?�� means low co-expression.