Refinement of a Quantitative Structure–Activity Relationship Model for Prediction of Cell-Penetrating Peptide Based Transfection Systems

Cell-penetrating peptide (CPP) based transfection systems (PBTS) are a promising class of drug delivery vectors. CPPs are short mainly cationic peptides capable of delivering cell non-permeant cargo to the interior of the cell. Some CPPs have the ability to form non-covalent complexes with oligonucleotides for gene therapy applications. In this study, we use quantitative structure–activity relationships (QSAR) , a statistical method based on regression data analysis. Here, an fragment QSAR (FQSAR) model is developed to predict new peptides based on standard alpha helical conformers and Assisted Model Building with Energy Refinement molecular mechanics simulations of previous peptides. These new peptides were examined for plasmid transfection efficiency and compared with their predicted biological activity. The best predicted peptides were capable of achieving plasmid transfection with significant improvement compared to the previous generation of peptides. Our results demonstrate that FQSAR model refinement is an efficient method for optimizing PBTS for improved biological activity.