临床价值导向下的AI制药范式转型:从FIC到FID的二元模型协同路径OA
Paradigm Shift in AI-Driven Drug Discovery under Clinical Value Orientation:A Synergistic Path of the Dual Model from FIC to FID
创新药产业是我国战略性支柱产业,《中华人民共和国药品管理法实施条例》明确以"临床价值为导向".以未满足临床需求为起点、患者获益为终点已成国际共识,但受限于疾病机制解析瓶颈,该理念长期未转化为可执行范式.本文整合人工智能(AI)技术体系,提出从"同类首创(FIC)"向"疾病领域治疗价值首创(FID)"的范式重构,核心在于将首创性判据由分子结构新颖性迁移至疾病治疗价值,论证 AI 与多组学如何使疾病机制解析及疾病优先逻辑走向可执行,并将其置于我国产业高质量发展语境中系统整合.本文沿"精准制造"与"科学决策"维度,将 AI 技术归并为侧重分子设计的化学模型与侧重机制解析的生物模型,此分类系契合研发全链条的功能性抽象.在此基础上,本文提出"二元模型携动是落实临床价值导向的关键路径",剖析因携动不足导致的研发与临床脱节痛点并提出多维建议,为产业高质量发展提供支撑.
The innovative drug industry is a strategic pillar industry in China.The Regulations for the Implementation of the Drug Administration Law explicitly emphasize a"clinical value-oriented"approach.Taking unmet clinical needs as the starting point and patient benefit as the endpoint has become an international consensus;however,constrained by bottlenecks in the elucidation of disease pathogenesis,this concept has long failed to be translated into an actionable paradigm.By integrating AI technologies,this paper proposes a paradigm shift from first-in-class(FIC)to first-in-disease(FID).The core of this reconstruction is to shift the criterion of pioneering status from molecular structural novelty to therapeutic value in disease treatment.It demonstrates how AI and multi-omics enable systematic dissection of disease pathogenesis and operationalize a disease-driven R&D logic and embed this new paradigm within China's framework for high-quality innovative drug industrial development.From the dimensions of"precision manufacturing"and"scientific decision-making,"this paper categorize AI-driven computational tools into two modules:chemistry models,which focus on molecular design,and biological models,which focus on the dissection of disease pathogenesis.This classification constitutes a functional abstraction aligned with the entire R&D chain.On this basis,the paper proposes that"the synergistic advancement of the dual models is the critical pathway to implementing the clinical value-oriented approach."It then analyzes the pain points of the R&D-clinical disconnect arising from insufficient synergy,and offers multidimensional recommendations to support the high-quality development of the industry.
赵宇;江旻;丛斌;王刚;张春明;张静;连云波;张堃
图灵-达尔文实验室||中科计算技术西部研究院北京大学肿瘤医院暨北京市肿瘤防治研究所国家药物临床试验机构,消化系肿瘤整合防治全国重点实验室河北医科大学首都医科大学附属北京安定医院中国科学院计算技术研究所北京哲源科技有限责任公司中国信息通信研究院中国科技产业投资管理有限公司
医药卫生
人工智能制药临床价值导向FICFID二元模型化学模型生物模型
artificial intelligence for drug discoveryclinical value orientationfirst-in-class(FIC)first-in-disease(FID)dual modelchemical modelbiological model
《中国食品药品监管》 2026 (6)
16-27,12
中国工程院院地合作项目(2025-DFZD-22)北京市科技计划项目(Z231100004623006)北京市自然科学基金-海淀原始创新联合基金(L232032)
评论