跳到正文

药代动力学与药效研究建模

组织 PK/PD 研究模型、参数、暴露与反应分析,保存数据依据、模型假设及评价结果。

开源精选医疗业务31 个文件50,135 Stars浏览次数 0K-Dense-AI/scientific-agent-skills MIT版本 a1043ef
开始使用

适用环境:Requires Python 3.12+ with NumPy 2+ and SciPy. No network access and no proprietary software. The estimation tools this skill orients you towards (NONMEM, Monolix, Phoenix, Simcyp, GastroPlus) are licensed separately and are never invoked by these scripts.

输入资料
获准使用的 PK/PD 研究资料 模型目标、参数与分析环境 阅读本节完整说明
交付内容
可复查的研究模型 模型评价与假设说明 阅读本节完整说明
  1. 获取完整包,保留配套文件的目录结构。
  2. 按照原文准备所需工具、依赖与输入资料。
  3. 阅读对应使用步骤与输入输出说明,在自己的环境中验证结果。

使用提示摘自原文,配套文件保留来源版本;实际运行条件以各包说明为准。

正文

获取原文件

references/regulatory-guidance.md

Reviewed 2026-10-01. ICH adoption and regional implementation are distinct; an EU effective date is not a worldwide effective date. Product-specific guidance and intended use can change the analysis. Check the applicable authority before submission.

Document Current verified boundary Primary source
ICH M12 Step 4 2024; EU effective 30 November 2024; enzyme/transporter DDI framework EMA
ICH M13A Immediate-release solid oral BE; EU effective 25 January 2025; transition leaves regional replicate/NTI provisions EMA
ICH M13B EMA Step 5 adopted 17 September 2026, effective 17 March 2027; additional-strength biowaivers Final PDF
ICH M13C Under development; do not describe as adopted harmonized HVD/NTI rules ICH workplan
ICH M15 General model-informed development; EU effective 23 July 2026; FDA final June 2026 EMA PDF, FDA
ICH E11A Pediatric extrapolation framework; exposure criteria justified for context FDA
FDA statistical BE Final May 2026; replaces 2001; Appendix G includes HVD replicate statistics FDA
FDA population PK Final February 2022; intended-use model development/evaluation FDA
E14/S7B Q&A 2022 integrated QT/proarrhythmia assessment; clinical/nonclinical evidence requirements FDA
FDA starting dose 2005 healthy-adult MRSD framework; not a universal MABEL calculator FDA

Apply the relevant question

  • NCA/BE: identify formulation, fasting/fed state, metric, sampling and jurisdiction; M13A is not a substitute for every special-design or product-specific requirement.
  • Population PK/ER: predefine the decision, data provenance, model assumptions, uncertainty and evaluation relevant to that use. Synthetic fit recovery is software evidence, not clinical evidence.
  • DDI: distinguish screening trigger from AUCR prediction and clinical action. Input uncertainty, metabolites and pathways outside the simplified model remain material.
  • Pediatrics/FIH: integrate biology, exposure-response, species/population relevance and safety; a numerical exposure match or conservative calculator output alone is insufficient.

No licensed estimation engine, prospective study, patient outcome or regulatory submission was validated by this skill refresh. The helper outputs do not certify compliance.

CONTACT HAONAN

联系haonan

医疗 AI · 产品与独立开发

刘浩南

关注医疗 AI 从 demo 到真实落地,聚焦产品化、商业化与院内工作流效率提升。

12年医疗产研经验,大湾区头部医疗AI公司前产品总监,独立开发者

医疗 AI 产品化医疗 AI 商业化院内工作流效率
刘浩南的微信联系二维码
微信扫一扫,联系 haonan