Proteochemometric Method for pIC50 Prediction of Flaviviridae

Abstract

Viruses remain an area of concern despite constant development of antiviral drugs and therapies. One of the contributors is the Flaviviridae family of viruses causing diseases that need attention. Among other anitviral methods, antiviral peptides are being studied as viable candidates. Although antiviral peptides (AVPs) are emerging as potential therapeutics, it is important to assess the efficacy of a given peptide in terms of its bioactivity. Experimental identification of the bioactivity of each potential peptide is an expensive and time consuming task. Computational methods like proteochemometric modeling (PCM) is a promising method for prediction of bioactivity (pIC50) based on peptide and target sequence pair. In this study, we propose a prediction of pIC50 of AVP against the Flaviviridae family that may help make the decision to choose a peptide with desired efficacy. The peptides data was collected from a public database and target sequences were manually curated from literature. Features are calculated using peptide and target sequence PCM descriptors which consist of individual and cross-term features of peptide and respective target. The resultant R2 and MAPE values are 0.85 and 8.44%, respectively, for prediction of pIC50 value of AVPs.