BMC Bioinformatics

Table 2 Q8 predictive precision of individual secondary structures from CB513

From: Prediction of 8-state protein secondary structures by a novel deep learning architecture

Q8 Label	CRRNN	NCCNN	MUFOLD-SS	DCRNN2^a	DCRNN	DeepCNF
H	0.86	0.841	0.855	0.863	0.832	0.849
B	0.466	0.676	0.571	0.571	0.554	0.433
E	0.797	0.767	0.764	0.768	0.753	0.748
G	0.466	0.487	0.413	0.419	0.429	0.49
I	0	0	0	0	0	0
T	0.556	0.577	0.572	0.562	0.559	0.53
S	0.494	0.548	0.522	0.509	0.518	0.487
L	0.603	0.565	0.586	0.571	0.573	0.571

^aData is generated by our experiment
Boldface numbers indicate best performance

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com