iRNAD

The original dataset composed of 550 RNA samples, of which each sequence samples is 41 nt long, which can be Downloaded.

Note that each sequence samples that we considered in this study is 23 nt long with at the center along the RNA sequences given below.

The benchmark dataset containing 550 RNA samples, of which 176 RNA sequence samples belong to the positive subset and 374 RNA sequence samples belong to the negative subset , which can be Downloaded.

Datasets of experiment I including the training dataset accounting for 80% of the whole dataset and the independent testing dataset accounting for 20% can be Downloaded.

Datasets of experiment II including the dataset only composed of sequence samples from S. cerevisiae, and the dataset consisting of the rest sequences from the other four species can be Downloaded.

The dataset from H. sapiens composed of 29 true D modification site sequences (positive dataset) and 68 false D modification site sequences (negative dataset) can be Downloaded.

The dataset from M. musculus composed of 13 true D modification site sequences (positive dataset) and 48 false D modification site sequences (negative dataset) can be Downloaded.

The dataset from D. melanogaster composed of 9 true D modification site sequences (positive dataset) and 38 false D modification site sequences (negative dataset) can be Downloaded.

The dataset from S. cerevisiae composed of 91 true D modification site sequences (positive dataset) and 93 false D modification site sequences (negative dataset) can be Downloaded.

The dataset from E.coli composed of 34 true D modification site sequences (positive dataset) and 127 false D modification site sequences (negative dataset) can be Downloaded.