RT-RICO: Protein Secondary Structure Prediction Using Rule Induction from Coverings

Note:

This webpage contains the PDB IDs for all protein sequences used (for training and testing) in research paper:

L. Lee, J. L. Leopold, R. L. Frank and A. M. Maglia, “Protein Secondary Structure Prediction Using Rule Induction from Coverings”, Proceedings of IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology 2009 (part of IEEE Symposium Series on Computational Intelligence 2009), Nashville, Tennessee, USA.

TABLE I, RESULTS FOR PROTEIN SECONDARY STRUCTURE PREDICTION (Click on links to view PDB IDs)

		Training Set
Folding Type Classes	Total Number of Proteins (SCOP)	Number of Proteins	Number of 5-Residue Segments	Number of Rules (at 90% threshold)
All-α	7,999	199	47,955	203,636
All-β	12,968	323	83,187	257,911
α/β	12,199	304	107,900	319,361
α+β	11,425	567	137,715	346,379

		Test Set
Folding Type Classes		Number of Proteins	Number of Residues	Q3 (%)
All-α		40	10,151	88.7
All-β		65	17,627	80.2
α/β		61	20,810	77.0
α+β		57	12,379	78.9

Total		223	60,967	80.3