斯坦福大学团队在《科学》期刊发表研究成果，利用生成式AI模型Evo1和Evo2设计出完整可复制的噬菌体基因组[1]。研究人员合成并测试了近300个生成的基因组，最终获得16种存活的噬菌体[1]，这些噬菌体混合物能够迅速击败对ΦX174具有抗性的大肠杆菌菌株[1]。
同时，由四川大学和华中科技大学联合开发的AI生物学家"荀子"在《自然-生物医学工程》发表论文[1]。该系统在2440万篇出版物和613.6 TB多源数据上进行训练[1]，可自主生成具有机制解释的疾病靶点假设，涵盖21008个人类基因和5850种疾病[1]。
此外，Moderna研发的mRNA流感疫苗mFLUSIVA获得美国FDA批准[1]，用于50岁及以上成人[1]，成为全球首款获批的季节性流感mRNA疫苗。该疫苗III期临床试验在11个国家纳入40805名50岁及以上成人[1]。
Researchers at Stanford University have successfully used generative AI models to design functional phage genomes, marking a significant advance in computational biology. On August 6th, the team's work was published in Science, describing how the Evo1 and Evo2 models generated approximately 300 phage genomes that were chemically synthesized and tested in the laboratory.[1] Of these, 16 viable phage variants were successfully created, and when combined together, the mixture rapidly overcame E. coli strains that had developed resistance to ΦX174.[1]
Academic teams from Sichuan University and Huazhong University of Science and Technology unveiled an AI biologist system called "Xunzi" on August 4th.[1] The system, published in Nature Biomedical Engineering, was trained on 24.4 million published scientific papers and 613.6 terabytes of multi-source data.[1] Xunzi covers 21,008 human genes and 5,850 diseases, and is capable of autonomously generating disease target hypotheses with mechanistic explanations.[1]
In related pharmaceutical news, Moderna's mRNA influenza vaccine mFLUSIVA received approval from the U.S. Food and Drug Administration on August 6th, becoming the first seasonal influenza mRNA vaccine to achieve regulatory clearance.[1] The vaccine is intended for adults aged 50 and older.[1] The Phase III clinical trial that supported this approval enrolled 40,805 participants aged 50 and above across 11 countries.[1]
