DMirNet: Inferring direct microRNA-mRNA association networks

BMC Systems Biology

Table 2 Performance comparison of DMirNet with the state-of-the-art Ensemble model

Dataset	Direct correlation inference methods					the state-of-the-art method
Dataset	Corpcor	Space	MIND	E-C&S&M	B&E-C&S&M	Pearson	IDA	Lasso	E-P&I&L
EMT	35	45	24	41	40	30	29	29	31
MCC	200	183	210	209	216	205	198	187	203
BR	98	83	95	102	105	114	124	120	101

To compare the performance of our method with a related work, we investigate the number of experimentally confirmed miRNA-mRNA associations of the state-of-the-art Ensemble model. It combines Pearson’s correlation (denoted as P), IDA (denoted as I), and Lasso (denoted as L) using the Borda count election and was reported as the best-performed Ensemble model on the three datasets [1]. ‘E’ denotes the Ensemble approach, and ‘B&E’ denotes the DMirNet with both bootstrapping and Ensemble aggregation

ISSN: 1752-0509