On 9 October 2024,in a high-profile vote of confidence for the promise of using artificial intelligence(AI)in scientific discovery,the Royal Swedish Academy of Sciences awarded Demis Hassabis(co-founder and chief exec...On 9 October 2024,in a high-profile vote of confidence for the promise of using artificial intelligence(AI)in scientific discovery,the Royal Swedish Academy of Sciences awarded Demis Hassabis(co-founder and chief executive officer)and John M.Jumper(direc-tor)of Google DeepMind(London,UK)the 2024 Nobel Prize in Chemistry for their pioneering work in developing the AI-powered protein structure prediction model AlphaFold2(AF2)[1].Also shar-ing the prize was David Baker(half to Hassabis and Jumper;half to Baker),professor of biochemistry at the University of Washington(Seattle,WA,USA),for his work on computational protein design that started with the mid-1990s development of Rosetta,a since-evolving suite of software tools that model protein structures using physical principles[2]-and now also AI[3].展开更多
AlphaFold3(AF3),as the latest generation of artificial intelligence model jointly developed by Google DeepMind and Isomorphic Labs,has been widely heralded in the scientific research community since its launch.With un...AlphaFold3(AF3),as the latest generation of artificial intelligence model jointly developed by Google DeepMind and Isomorphic Labs,has been widely heralded in the scientific research community since its launch.With unprecedented accuracy,the AF3 model may successfully predict the structure and interactions of virtually all biomolecules,including proteins,ligands,nucleic acids,ions,etc.By accurately simulating the structural information and interactions of biomacromolecules,it has shown great potential in many aspects of structural prediction,mechanism research,drug design,protein engineering,vaccine development,and precision therapy.In order to further understand the characteristics of AF3 and accelerate its promotion,this article sets out to address the development process,working principle,and application in drugs and biomedicine,especially focusing on the intricate differences and some potential pitfalls compared to other deep learning models.We explain how a structure-prediction tool can impact many research fields,and in particular revolutionize the strategies for designing of effective next generation vaccines and chemical and biological drugs.展开更多
Interactions between macromolecules orchestrate many mechanobiology processes.However,progress in the field has often been hindered by the monetary and time costs of obtaining reliable experimental structures.In recen...Interactions between macromolecules orchestrate many mechanobiology processes.However,progress in the field has often been hindered by the monetary and time costs of obtaining reliable experimental structures.In recent years,deep-learning methods,such as AlphaFold,have democratized access to high-quality predictions of the structural properties of proteins and other macromolecules.The newest implementation,AlphaFold 3,significantly expands the applications of its predecessor,AlphaFold 2,by incorporating reliable models for small molecules and nucleic acids and enhancing the prediction of macromolecular complexes.While several limitations still exist,the continuous improvement of machine learning methods like AlphaFold is producing a significant revolution in the field.The possibility of easily accessing structural predictions of biomolecular complexes may create substantial impacts in mechanobiology.Indeed,structural studies are at the basis of several applications in the field,such as drug discovery for mechanosensing proteins,development of mechanotherapy,understanding the mechanotransduction mechanisms and the mechanistic basis of diseases,or designing biomaterials for tissue engineering.展开更多
文摘On 9 October 2024,in a high-profile vote of confidence for the promise of using artificial intelligence(AI)in scientific discovery,the Royal Swedish Academy of Sciences awarded Demis Hassabis(co-founder and chief executive officer)and John M.Jumper(direc-tor)of Google DeepMind(London,UK)the 2024 Nobel Prize in Chemistry for their pioneering work in developing the AI-powered protein structure prediction model AlphaFold2(AF2)[1].Also shar-ing the prize was David Baker(half to Hassabis and Jumper;half to Baker),professor of biochemistry at the University of Washington(Seattle,WA,USA),for his work on computational protein design that started with the mid-1990s development of Rosetta,a since-evolving suite of software tools that model protein structures using physical principles[2]-and now also AI[3].
基金supported by funds of the Key Research and Development Grant of MOST(grant No.2023YFA095000)the Affiliated Xiangshan Hospital of Wenzhou Medical University,the National Science Foundation of China(grant No.82470005)Wenzhou Institute University of Chinese Academy of Sciences.
文摘AlphaFold3(AF3),as the latest generation of artificial intelligence model jointly developed by Google DeepMind and Isomorphic Labs,has been widely heralded in the scientific research community since its launch.With unprecedented accuracy,the AF3 model may successfully predict the structure and interactions of virtually all biomolecules,including proteins,ligands,nucleic acids,ions,etc.By accurately simulating the structural information and interactions of biomacromolecules,it has shown great potential in many aspects of structural prediction,mechanism research,drug design,protein engineering,vaccine development,and precision therapy.In order to further understand the characteristics of AF3 and accelerate its promotion,this article sets out to address the development process,working principle,and application in drugs and biomedicine,especially focusing on the intricate differences and some potential pitfalls compared to other deep learning models.We explain how a structure-prediction tool can impact many research fields,and in particular revolutionize the strategies for designing of effective next generation vaccines and chemical and biological drugs.
基金supported by Xi'an Jiaotong-Liverpool University(Research Development Fund RDF-23-01-026 to F.Z)FOCEM(MERCOSUR Structural Convergence Fund-COF 03/11)to S.P.
文摘Interactions between macromolecules orchestrate many mechanobiology processes.However,progress in the field has often been hindered by the monetary and time costs of obtaining reliable experimental structures.In recent years,deep-learning methods,such as AlphaFold,have democratized access to high-quality predictions of the structural properties of proteins and other macromolecules.The newest implementation,AlphaFold 3,significantly expands the applications of its predecessor,AlphaFold 2,by incorporating reliable models for small molecules and nucleic acids and enhancing the prediction of macromolecular complexes.While several limitations still exist,the continuous improvement of machine learning methods like AlphaFold is producing a significant revolution in the field.The possibility of easily accessing structural predictions of biomolecular complexes may create substantial impacts in mechanobiology.Indeed,structural studies are at the basis of several applications in the field,such as drug discovery for mechanosensing proteins,development of mechanotherapy,understanding the mechanotransduction mechanisms and the mechanistic basis of diseases,or designing biomaterials for tissue engineering.
