Table 2.

Performance of PETcofold and other joint structure prediction methods on four sRNA–mRNA examples

sRNA-target pair	MCC										Run time (s)
	PETcofold		inteRNA		Pairfold		RactIP		RNAcofold		PETcofold		inteRNA	Pairfold	RactIP	RNAcofold
	-extstem		80%	100%	80%	100%	80%	100%	80%	100%	-extstem
MicA-ompA	0.87	0.83	0.49	0.51	0.86	0.74	0.57	0.57	0.80	0.67	28.7	28.4	69493.1	3.2	3.0	0.2
OxyS-fhlA	0.80	0.82	0.64	0.64	0.61	0.61	0.48	0.48	0.61	0.61	20.6	19.3	129636.7	1.9	2.0	0.2
RyhB-uof-fur	0.13	0.13	0.12	0.00	0.21	0.21	0.19	0.00	0.21	0.21	26.4	25.3	65599.2	2.6	2.7	0.2
RyhB-sodB	0.67	0.71	0.70	0.68	0.65	0.51	0.65	0.59	0.65	0.63	15.4	15.2	23579.3	1.7	2.0	0.1
Average	0.62	0.62	0.49	0.46	0.58	0.52	0.47	0.41	0.57	0.53	22.8	22.1	72077.1	2.4	2.5	0.2

The MCC evaluates the joint structure, i.e. both the interaction between the two RNAs and the secondary structure of each single RNA. PETcofold was called with parameters δ = 0.9, γ = 0.1, option -noLP and optionally with option -extstem. inteRNA, Pairfold and RactIP were called with default parameters. RNAcofold was called with options -d2 -noLP. The columns 80% and 100% give the result for the consensus structure with base pairs that occur in 80 and 100%, respectively, of the single structures. The run time of all single sequence-based approaches is the sum for all input sequences.