Skip to main content

Table 2 Cross-validated density prediction accuracy of NBDP models. Average KL divergence for the Naive Bayes' Density Predictor is shown for the 3- and 8-class prediction tasks with varying sequence window sizes. The residues in the window were described using the amino acid identity method. All predictions are for 10-fold cross-validation on the training set (set-174). Best results are shown in bold.

From: Prediction of protein continuum secondary structure with probabilistic models based on NMR solved structures

window size 3-class accuracy 8-class accuracy
5 0.77 1.20
7 0.75 1.19
9 0.75 1.19
11 0.74 1.20
13 0.75 1.22
15 0.76 1.25
17 0.78 1.28
19 0.79 1.31