Moderna President Says AI Can Accelerate Personalized Drug Discovery
Moderna President Stephen Hoge said AI already assists with mRNA sequence and protein design, clinical-trial design, and patient recruitment, though he described these as company claims. He said Moderna’s personalized cancer-vaccine workflow can incorporate information from up to 34 new antigens, use as many as nine doses, and take about six weeks from receiving a qualified sample to returning the drug to a hospital. Hoge cautioned that treatment efficacy and durability still require human studies, and that faster discovery does not accelerate manufacturing, recruitment, regulatory approval, or clinical validation.
The remarks show how AI is already being used in one drug developer’s research and personalized-vaccine workflow, while highlighting that clinical benefits and broad efficiency gains remain unvalidated.
在 10 月 20 日上线的黑石播客《Inside Blackstone》中，主持人问起 Moderna 总裁斯蒂芬·霍格（Stephen Hoge），AI 能否在未来三五年带来改变治疗的新突破时，霍格这样回答，他相信 AI 能够加快药物发现，但也提醒，即使把这个过程缩短到几个月，甚至几周，研究者仍然需要在人身上观察疗效，才能判断一种药物是否有效、效果能持续多久。
新冠疫苗的速度是怎样积累出来的？
霍格的判断来源于他自身的经历。他在 2012 年加入 Moderna 时，公司还只有约二十个人。在此之前，他做过急诊科医生，又在麦肯锡从事了七年咨询工作。让他决定加入这家初创公司的是 mRNA 作为药物的可能性：mRNA 在人体内承担着传递蛋白质制造指令的工作；如果能够把经过设计的 mRNA 送入细胞，就有机会让细胞生产预防或治疗疾病所需的蛋白。沿着这个思路，研究者可以尝试更换指令，开发针对其他疾病的药物。