Four measurements were used to evaluate the prediction performances of the proposed model: accuracy (Ac), sensitivity (Sn), specificity (Sp) and Matthews' correlation coefficient (MCC).
We used sensitivity (S n ), specificity (S p ), accuracy (ACC) and Matthew's Correlation Coefficient (MCC) [ 33] as the measurements to evaluate the prediction performance of the classifiers.
In a classification problem, the specificity (Sp), sensitivity (Sn), accuracy (Ac) and Mathew's correlation coefficient (MCC) are most widely used to evaluate the prediction system.
The particular allotypes used for testing were chosen for two reasons: (1) adequate peptide binding data was available for evaluating the prediction results and (2) they have very different peptide binding specificities so that the results reflect the generality of the prediction method for multiple MHC allotypes.
We evaluated the prediction performance of netSVM using ROC analysis, from which AUC, accuracy, sensitivity, and specificity were calculated.
We evaluated the predictions on each sample.
In order to evaluate the specificity of motif prediction, we compared the SH2 selectivity values (which is calculated using enrichment values from peptide library screening) of SLiMs in proteins from reported binding groups to the SH2 selectivity values of SLiMs in proteins from groups that are not reported to bind.
The ROC analysis is a standard approach to evaluate the sensitivity and specificity of prediction methods.
Otherwise, it is counted as a false-positive prediction when we evaluate the specificity of our method (see Section 3.2).
The miR-mRNA pairs from the interactions set are reported in a binary format, as being either functional or non-functional, and we use the area under the ROC curve (AUC) measure to evaluate the sensitivity and specificity of our prediction method.
