Main Article Content
Antimicrobial peptides (AMPs) have been considered as potential therapeutic agents against drug-resistant bacterial pathogens. Accurately relating the quantitative properties of AMPs with the activity may expedite novel AMP design. In this study, the structural and physicochemical features of peptides computed by the updated PROFEAT web server were applied to develop QSAR models for 101cationic peptides (CAMEL-s). The variables were optimized by stepwise multiple regression (SMR) and genetic algorithm (GA), QSAR models were constructed by multiple linear regression (MLR) and partial least squares (PLS), respectively. These models show good performance in reliability and predictability. The results indicate that these peptide descriptors may be promising in QSAR research of AMPs, and the QSAR models should be useful for novel AMP design.