Classification of protein sequences by means of irredundant patterns

BMC Bioinformatics

Table 5 Comparison of results against state-of-the-art methods

Learning algorithm	Mean ROC	Mean ROC50	Mean mRFP
Irredundant Class	0.929	0.524	0.0554
Local Alignment ("ekm," β = 0.5)	0.929	0.600	0.0515
Local Alignment ("eig," β = 0.5)	0.925	0.649	0.0541
Word Correlation Matrices (k = 6)	0.904	0.447	0.0778
Pairwise	0.896	0.464	0.0837
Mismatch (k = 5, m = 1)	0.872	0.400	0.0837
Spectrum (k = 3)	0.824	0.294	0.1535
Fisher	0.773	0.250	0.2040

The comparison is based on mean scores of ROC, ROC50 (ROC curve up to the first 50 false positives), and Median rate of false positives (mRFP) for the Irredundant Class and state-of-the-art methods. In bold are reported the best results for each score.

ISSN: 1471-2105