期刊
VIRUSES-BASEL
卷 14, 期 3, 页码 -出版社
MDPI
DOI: 10.3390/v14030465
关键词
SARS-CoV-2; spike-protein; delta variant; interatomic interaction; amino-acid-amino-acid bond pair; machine learning
类别
资金
- National Science Foundation of USA [RAPID DMR/CMMT2028803, RAPID CISE/CNS2034247]
The SARS-CoV-2 Delta variant, with its rapid spread and high infection rate, has a mutation that allows the virus to invade human cells more efficiently. Research shows that the mutation reduces the bonding cohesion between interacting residues, increasing flexibility and causing damage to the virus. The mutation quantifiers could potentially be used in machine learning protocols to predict emerging mutations.
The SARS-CoV-2 Delta variant is emerging as a globally dominant strain. Its rapid spread and high infection rate are attributed to a mutation in the spike protein of SARS-CoV-2 allowing for the virus to invade human cells much faster and with an increased efficiency. In particular, an especially dangerous mutation P681R close to the furin cleavage site has been identified as responsible for increasing the infection rate. Together with the earlier reported mutation D614G in the same domain, it offers an excellent instance to investigate the nature of mutations and how they affect the interatomic interactions in the spike protein. Here, using ultra large-scale ab initio computational modeling, we study the P681R and D614G mutations in the SD2-FP domain, including the effect of double mutation, and compare the results with the wild type. We have recently developed a method of calculating the amino-acid-amino-acid bond pairs (AABP) to quantitatively characterize the details of the interatomic interactions, enabling us to explain the nature of mutation at the atomic resolution. Our most significant finding is that the mutations reduce the AABP value, implying a reduced bonding cohesion between interacting residues and increasing the flexibility of these amino acids to cause the damage. The possibility of using this unique mutation quantifiers in a machine learning protocol could lead to the prediction of emerging mutations.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据