Ab Initio Design of Amphipathic-Symmetric Peptides against SARS-CoV-2

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes severe health crisis and huge socioeconomic upheaval internationally. This study proposes an ab initio design strategy to obtain antiviral peptides to against SARS-CoV-2 infection. The study employed database filtering technology to generate 7 amphipathic-symmetric peptides named DFTavPs with low cytotoxicity and random coil structure. Three DFTavPs promoted SARS-CoV-2 pseudoviruses infection and three DFTavPs inhibited virus infection, which are accompanied by up-regulation or down-regulation of SARS-CoV receptor angiotensin-converting enzyme 2 mRNA levels. Particularly, microRNA profiling showed that some differentially expressed microRNAs had potential to target key factors for cell entry of SARS-CoV-2. Furthermore, we explored the relationship of parameters and antiviral efficacy index (AEI). The results suggested that higher AEI of coronavirus was most likely to occur at mean amphipathic moment between 0.3 and 0.4. Automated machine learning was used to construct parameters-AEI regression models for various viruses. The Extra-Trees and CatBoost had a good predicting performance for AEI of coronavirus (R2=0.794 and Rpearson=0.897) and human immunodeficiency virus (R2=0.735 and Rpearson=0.859), respectively. Overall, this strategy is expected to efficiently obtain huge amounts of potential peptide drugs with anti-SARS-CoV-2 activity, and machine learning models could contribute to discovery of high antivirus-activity peptides. Examination of 2 antiviral peptides influence in 293T cells.