PEP-PREDNa+: A web server for prediction of highly specific peptides targeting voltage-gated Na+ channels using machine learning techniques

Abstract

Voltage-gated sodium channel activity has long been associated with several diseases including epilepsy, chronic pain, cardiovascular diseases, cancers, immune system, neuromuscular and respiratory disorders. The strong participation of these channels in the development of diseases makes them excellent promising therapeutic targets. Voltage-gated Na+ channel blocking peptides come from a wide source of organisms such as venoms. However, the in vitro and in vivo identification and validation of these peptides are time-consuming and resource-intensive. In this work, we developed a bioinformatics tool called PEP-PREDNa + for the highly specific prediction of voltage-gated Na+ channel blocking peptides. PEP-PREDNa+ is based on the random forest algorithm, which presented excellent performance measures during the cross-validation (sensitivity = 0.81, accuracy = 0.83, precision = 0.85, F-score = 0.83, specificity = 0.86, and Matthew's correlation coefficient = 0.67) and testing (sensitivity = 0.88, accuracy = 0.92, precision = 0.96, F-score = 0.91, specificity = 0.96, and Matthew's correlation coefficient = 0.84) phases. The PEP-PREDNa + tool could be very useful in accelerating and reducing the costs of the discovery of new voltage-gated Na+ channel blocking peptides with therapeutic potential.