Systematic exploration of autonomous modules in noisy microRNA-target networks for testing the generality of the ceRNA hypothesis

Table 1 Summary statistics of the different datasets used in our study

		High confidence			High coverage
Dataset	Type	miRNAs	mRNAs	Interactions	miRNAs	mRNAs	Interactions
TarBase [21]	Validated	202	2,315	9,569	202	2,315	9,569
miRanda [6]	Predicted	409	567	1,000	856	2,633	5,000
miRGen [68]	Predicted	503	37	1,041	828	103	5,230
PicTar [69]	Predicted	91	438	1,000	164	1,541	5,000
PITA [70]	Predicted	290	760	1,000	582	2,708	5,000
TargetScan [71]	Predicted	29	214	1,000	40	948	5,000
Union (without TarBase)	Integrated	926	1,818	4,983	1,505	6,063	24,553
Expressed union (without TarBase)	Integrated	163	448	701	240	2,034	4,337
Union (with TarBase)	Integrated	1,063	3,711	14,548	1,631	7,208	34,111
Expressed union (with TarBase)	Integrated	181	605	1,020	256	2,188	4,653

The integrated datasets involve data from all the prediction sets with or without the experimentally validated miRNA-target pairs. Among them, the expressed union sets were formed by considering only the expressed miRNAs and mRNAs in the corresponding union sets. We used these four expressed union sets (with or without TarBase, high confidence or high coverage) in our analyses.

ISSN: 1471-2164