TaxDiff:taxonomic-guided diffusion model for protein sequence generation  

作　　者：Zongying LIN Hao LI Liuzhenghao LV Yu WANG Bin LIN Junwu ZHANG Zijun CHEN Calvin Yu-Chian CHEN Li YUAN Yonghong TIAN 

机构地区：[1]School of Electronic and Computer Engineering,Peking University,Shenzhen 518055,China [2]Peng Cheng Laboratory,Shenzhen 518000,China [3]AI for Science(AI4S)-Preferred Program,Peking University Shenzhen Graduate School,Shenzhen 518055,China [4]School of Chemical Biology and Biotechnology,Peking University,Shenzhen 518055,China

出　　处：《Science China(Information Sciences)》2025年第4期381-382,共2页中国科学(信息科学)(英文版)

基　　金：supported in part by National Natural Science Foundation of China(Grant Nos.62425101,62088102,62202014);Shenzhen Science and Technology Program(Grant No.KQTD20240729102051063)。

摘　　要：Protein design aims to generate protein variants with targeted biological functions,which is significant in multiple biological areas,including enzyme reaction catalysis,vaccine design,and fluorescence intensity.Protein design contains two paradigms:sequence generation and structure generation.Recently,EvoDiff[1]proposed a universal designing paradigm,combining structure and sequence generation using the diffusion framework,which improves the protein design efficiency.

关 键 词：enzyme reaction catalysisvaccine designand generate protein variants targeted biological functionswhich structure generationrecentlyevodiff proposed structure sequence generation fluorescence intensityprotein design diffusion frameworkwhich sequence generation protein design 

分 类 号：Q51[生物学—生物化学]