The naive Bayesian scoring scheme combines the histogram-based learning method with Bayesian rules to estimate a probability for assigning a spin system to an amino acid residue residing in a specific second structure. The normal distribution score scheme assumes that the chemical shifts contained in the spin system follows the Gaussian distribution and use the normal density function to estimate the probability.

The non_probabilistic score option makes scoring schemes only use the second structure type information in the PsiPred output. The probabilistic score option makes scoring schemes use the probability information for each predicted secondary structure type in the PSIPRED output.