iCircDA-LTR:

Dataset

The circRNA-disease association dataset is divided into different sub-datasets according to two application scenarios.

(1) 𝕊all+	: 4195 experimentally validated circRNA-disease associations (All-positive-associations.txt)

For the first application scenario 'detection of missing associations', there is no overlapping circRNA-disease association between the independent test dataset and benchmark dataset:

(2) 𝕊benchmark1+	: 3356 positive circRNA-disease associations (Benchmark-pos-1.txt)
(3) 𝕊benchmark1-	: 263332 negative circRNA-disease associations (Benchmark-neg-1.txt)
(4) 𝕊independent1+	: 839 positive circRNA-disease associations (Independent-pos-1.txt)
(5) 𝕊independent1-	: 65833 negative circRNA-disease associations (Independent-neg-1.txt)

For the second application scenario 'prediction of disease associated with newly discovered circRNAs', there is no overlapping circRNA between the independent test dataset and benchmark dataset:

(6) 𝕊benchmark2+	: 3365 positive circRNA-disease associations (Benchmark-pos-2.txt)
(7) 𝕊benchmark2-	: 263395 negative circRNA-disease associations (Benchmark-neg-2.txt)
(8) 𝕊independent2+	: 830 positive circRNA-disease associations (Independent-pos-2.txt)
(9) 𝕊independent2-	: 65770 negative circRNA-disease associations (Independent-neg-2.txt)