February 12, 2024 | Chentong Wang, Yannan Qu, Zhangzhi Peng, Yukai Wang, Hongli Zhu, Dachuan Chen, Longxing Cao
Proteus is an innovative approach to protein structure generation that aims to enhance designability and efficiency in *de novo* protein design. It leverages graph-based triangle methods and a multi-track interaction network to generate designable protein structures. Proteus demonstrates superior performance in computational evaluations, achieving high designability and efficiency. The model's reliability is validated through experimental characterization, showing a high success rate in generating designable proteins. Proteus's ability to rapidly create highly designable protein backbones without pre-training techniques significantly advances our understanding of protein structure diffusion and contributes to advancements in protein design. The source code of Proteus will be released once accepted.Proteus is an innovative approach to protein structure generation that aims to enhance designability and efficiency in *de novo* protein design. It leverages graph-based triangle methods and a multi-track interaction network to generate designable protein structures. Proteus demonstrates superior performance in computational evaluations, achieving high designability and efficiency. The model's reliability is validated through experimental characterization, showing a high success rate in generating designable proteins. Proteus's ability to rapidly create highly designable protein backbones without pre-training techniques significantly advances our understanding of protein structure diffusion and contributes to advancements in protein design. The source code of Proteus will be released once accepted.