Table 2

Cross-validation performance of the predictive models trained with various features.

Training features

Pre

Sn

Sp

Acc

MCC

Positional Weighted Matrix of flanking Amino Acids (AA_PWM)

0.735

0.817

0.843

0.834

0.646

Amino Acid Composition (AAC)

0.696

0.798

0.814

0.808

0.596

Accessible Surface Area (ASA)

0.672

0.768

0.800

0.789

0.553

Secondary structure (SS)

0.580

0.718

0.723

0.721

0.424

AA_PWM + AAC

0.738

0.814

0.846

0.835

0.647

AA_PWM + ASA

0.781

0.831

0.876

0.860

0.698

AA_PWM + SS

0.709

0.791

0.827

0.814

0.604

AA_PWM + AAC + ASA

0.836

0.860

0.910

0.892

0.765

AA_PWM + AAC + SS

0.711

0.801

0.827

0.818

0.613

AA_PWM + AAC + ASA + SS

0.812

0.860

0.894

0.882

0.745

Abbreviation: Pre, precision; Sn, sensitivity; Sp, specificity; Acc, accuracy; MCC, Matthews Correlation Coefficient.

Lee et al. BMC Bioinformatics 2011 12(Suppl 13):S10   doi:10.1186/1471-2105-12-S13-S10