Table 1.

SEN, PPV, MCC and F-score for each prediction algorithm

	McCaskill model				CONTRAfold model
γ	SEN	PPV	MCC	F-score	SEN	PPV	MCC	F-score
gc-pMCC	0.67	0.54	0.60	0.60	0.68	0.69	0.68	0.68
gc-pF	0.67	0.54	0.60	0.59	0.69	0.68	0.68	0.69
0.03125	0.34	0.78	0.52	0.48	0.09	0.90	0.28	0.16
0.0625	0.40	0.74	0.54	0.52	0.14	0.90	0.36	0.24
0.125	0.45	0.69	0.56	0.55	0.21	0.86	0.42	0.34
0.25	0.51	0.66	0.58	0.58	0.30	0.82	0.50	0.44
0.5	0.57	0.62	0.59	0.60	0.41	0.79	0.56	0.54
1.0	0.65	0.58	0.61	0.61	0.53	0.75	0.63	0.62
2.0	0.67	0.53	0.60	0.60	0.65	0.70	0.67	0.67
4.0	0.68	0.51	0.59	0.59	0.72	0.65	0.68	0.68
6.0	0.69	0.51	0.59	0.58	0.74	0.62	0.68	0.68
8.0	0.69	0.50	0.59	0.58	0.75	0.61	0.68	0.67
16.0	0.69	0.49	0.58	0.57	0.76	0.58	0.66	0.66
32.0	0.69	0.48	0.58	0.57	0.76	0.56	0.65	0.64
64.0	0.69	0.48	0.57	0.56	0.76	0.54	0.64	0.64
128.0	0.69	0.47	0.57	0.56	0.76	0.53	0.64	0.63
256.0	0.69	0.47	0.57	0.56	0.76	0.53	0.63	0.62
512.0	0.69	0.47	0.57	0.56	0.76	0.52	0.63	0.62
1024.0	0.69	0.47	0.57	0.56	0.76	0.52	0.63	0.62
RNAfold	0.65	0.50	0.57	0.57	-
Simfold	0.64	0.51	0.57	0.57	-
Sfold	0.65	0.58	0.62	0.62	-

The rows labeled "gc-pMCC" and "gc-pF" indicate RNA secondary structure prediction with the γ-centroid estimator and the pseudo-expected MCC and F-score, respectively (Method M2). The rows below the dashed line indicate the results of the γ-centroid estimator [10] with various values of the γ parameter (given in the first column). Note that Simfold and Sfold employ a similar probabilistic model (based on energy parameters) for secondary structures to the McCaskill model.