• 文献检索
  • 文档翻译
  • 深度研究
  • 学术资讯
  • Suppr Zotero 插件Zotero 插件
  • 邀请有礼
  • 套餐&价格
  • 历史记录
应用&插件
Suppr Zotero 插件Zotero 插件浏览器插件Mac 客户端Windows 客户端微信小程序
定价
高级版会员购买积分包购买API积分包
服务
文献检索文档翻译深度研究API 文档MCP 服务
关于我们
关于 Suppr公司介绍联系我们用户协议隐私条款
关注我们

Suppr 超能文献

核心技术专利:CN118964589B侵权必究
粤ICP备2023148730 号-1Suppr @ 2026

文献检索

告别复杂PubMed语法,用中文像聊天一样搜索,搜遍4000万医学文献。AI智能推荐,让科研检索更轻松。

立即免费搜索

文件翻译

保留排版,准确专业,支持PDF/Word/PPT等文件格式,支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述,25分钟生成高质量综述,智能提取关键信息,辅助科研写作。

立即免费体验

条件通用微分方程描述了C肽产生中的群体动态和个体间差异。

Conditional universal differential equations capture population dynamics and interindividual variation in c-peptide production.

作者信息

de Rooij Max, van Riel Natal A W, O'Donovan Shauna D

机构信息

Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands.

Eindhoven Artificial Intelligence Systems Institute (EAISI), Eindhoven University of Technology, Eindhoven, The Netherlands.

出版信息

NPJ Syst Biol Appl. 2025 Jul 31;11(1):84. doi: 10.1038/s41540-025-00570-6.

DOI:10.1038/s41540-025-00570-6
PMID:40744933
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC12313987/
Abstract

Universal differential equations (UDEs) are an emerging approach in biomedical systems biology, integrating physiology-driven mathematical models with machine learning for data-driven model discovery in areas where knowledge of the underlying physiology is limited. However, current approaches to training UDEs do not directly accommodate heterogeneity in the underlying data. As a data-driven approach, UDEs are also vulnerable to overfitting and consequently cannot sufficiently generalize to heterogeneous populations. We propose a conditional UDE (cUDE) where we assume that the structure and weights of the embedded neural network are common across individuals, and introduce a conditioning parameter that is allowed to vary between individuals. In this way, the cUDE architecture can accommodate inter-individual variation in data while learning a generalizable network representation. We demonstrate the effectiveness of the cUDE as an extension of the UDE framework by training a cUDE model of c-peptide production. We show that our cUDE model can accurately describe postprandial c-peptide levels in individuals with normal glucose tolerance, impaired glucose tolerance, and type 2 diabetes mellitus. Furthermore, we show that the conditional parameter captures relevant inter-individual variation. Subsequently, we use symbolic regression to derive a generalizable analytical expression for c-peptide production.

摘要

通用微分方程(UDEs)是生物医学系统生物学中一种新兴的方法,它将生理学驱动的数学模型与机器学习相结合,用于在基础生理学知识有限的领域进行数据驱动的模型发现。然而,当前训练UDEs的方法并不能直接适应基础数据中的异质性。作为一种数据驱动的方法,UDEs也容易出现过拟合,因此不能充分推广到异质群体。我们提出了一种条件UDE(cUDE),我们假设嵌入神经网络的结构和权重在个体之间是通用的,并引入一个允许在个体之间变化的条件参数。通过这种方式,cUDE架构可以在学习可推广的网络表示的同时适应数据中的个体间差异。我们通过训练一个c肽产生的cUDE模型,证明了cUDE作为UDE框架扩展的有效性。我们表明,我们的cUDE模型可以准确描述糖耐量正常、糖耐量受损和2型糖尿病个体的餐后c肽水平。此外,我们表明条件参数捕获了相关的个体间差异。随后,我们使用符号回归来推导c肽产生的可推广解析表达式。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/14acf027ca6b/41540_2025_570_Fig6_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/8781fd6b6fc8/41540_2025_570_Fig1_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/bc49a8f929b8/41540_2025_570_Fig2_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/e4439ba1083d/41540_2025_570_Fig3_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/ab8cbd305813/41540_2025_570_Fig4_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/917cbf6e2af4/41540_2025_570_Fig5_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/14acf027ca6b/41540_2025_570_Fig6_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/8781fd6b6fc8/41540_2025_570_Fig1_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/bc49a8f929b8/41540_2025_570_Fig2_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/e4439ba1083d/41540_2025_570_Fig3_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/ab8cbd305813/41540_2025_570_Fig4_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/917cbf6e2af4/41540_2025_570_Fig5_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/575e/12313987/14acf027ca6b/41540_2025_570_Fig6_HTML.jpg

相似文献

1
Conditional universal differential equations capture population dynamics and interindividual variation in c-peptide production.条件通用微分方程描述了C肽产生中的群体动态和个体间差异。
NPJ Syst Biol Appl. 2025 Jul 31;11(1):84. doi: 10.1038/s41540-025-00570-6.
2
Short-Term Memory Impairment短期记忆障碍
3
The health economics of insulin therapy: How do we address the rising demands, costs, inequalities and barriers to achieving optimal outcomes.胰岛素治疗的卫生经济学:我们如何应对不断增长的需求、成本、不平等现象以及实现最佳治疗效果的障碍。
Diabetes Obes Metab. 2025 Jul;27 Suppl 5(Suppl 5):24-35. doi: 10.1111/dom.16488. Epub 2025 Jun 4.
4
Preexisting Diabetes and Pregnancy: An Endocrine Society and European Society of Endocrinology Joint Clinical Practice Guideline.孕前糖尿病与妊娠:内分泌学会和欧洲内分泌学会联合临床实践指南
Eur J Endocrinol. 2025 Jun 30;193(1):G1-G48. doi: 10.1093/ejendo/lvaf116.
5
Multiscale Probabilistic Modeling: A Bayesian Approach to Augment Mechanistic Models of Cell Signaling with Machine-Learning Predictions of Binding Affinity.多尺度概率建模:一种利用结合亲和力的机器学习预测增强细胞信号传导机制模型的贝叶斯方法。
bioRxiv. 2025 Jul 9:2025.05.23.655795. doi: 10.1101/2025.05.23.655795.
6
Preexisting Diabetes and Pregnancy: An Endocrine Society and European Society of Endocrinology Joint Clinical Practice Guideline.糖尿病合并妊娠:内分泌学会与欧洲内分泌学会联合临床实践指南
J Clin Endocrinol Metab. 2025 Jul 13. doi: 10.1210/clinem/dgaf288.
7
Sexual Harassment and Prevention Training性骚扰与预防培训
8
Systemic Inflammatory Response Syndrome全身炎症反应综合征
9
The quantity, quality and findings of network meta-analyses evaluating the effectiveness of GLP-1 RAs for weight loss: a scoping review.评估胰高血糖素样肽-1受体激动剂(GLP-1 RAs)减肥效果的网状Meta分析的数量、质量及结果:一项范围综述
Health Technol Assess. 2025 Jun 25:1-73. doi: 10.3310/SKHT8119.
10
Comparison of Two Modern Survival Prediction Tools, SORG-MLA and METSSS, in Patients With Symptomatic Long-bone Metastases Who Underwent Local Treatment With Surgery Followed by Radiotherapy and With Radiotherapy Alone.两种现代生存预测工具 SORG-MLA 和 METSSS 在接受手术联合放疗和单纯放疗治疗有症状长骨转移患者中的比较。
Clin Orthop Relat Res. 2024 Dec 1;482(12):2193-2208. doi: 10.1097/CORR.0000000000003185. Epub 2024 Jul 23.

本文引用的文献

1
Assessment of uncertainty quantification in universal differential equations.通用微分方程中不确定性量化的评估
Philos Trans A Math Phys Eng Sci. 2025 Apr 2;383(2293):20240444. doi: 10.1098/rsta.2024.0444.
2
Physiology-informed regularisation enables training of universal differential equation systems for biological applications.基于生理学的正则化能够训练用于生物学应用的通用微分方程系统。
PLoS Comput Biol. 2025 Jan 23;21(1):e1012198. doi: 10.1371/journal.pcbi.1012198. eCollection 2025 Jan.
3
Subclassification of obesity for precision prediction of cardiometabolic diseases.
肥胖的亚分类用于心血管代谢疾病的精准预测。
Nat Med. 2025 Feb;31(2):534-543. doi: 10.1038/s41591-024-03299-7. Epub 2024 Oct 24.
4
Does GLP-1 cause post-bariatric hypoglycemia: 'Computer says no'.GLP-1 是否会引起减重手术后的低血糖:“电脑说不”。
Comput Methods Programs Biomed. 2024 Dec;257:108424. doi: 10.1016/j.cmpb.2024.108424. Epub 2024 Sep 19.
5
Quantifying the effect of nutritional interventions on metabolic resilience using personalized computational models.使用个性化计算模型量化营养干预对代谢恢复力的影响。
iScience. 2024 Feb 29;27(4):109362. doi: 10.1016/j.isci.2024.109362. eCollection 2024 Apr 19.
6
Estimating insulin sensitivity and β-cell function from the oral glucose tolerance test: validation of a new insulin sensitivity and secretion (ISS) model.从口服葡萄糖耐量试验估计胰岛素敏感性和β细胞功能:新的胰岛素敏感性和分泌(ISS)模型的验证。
Am J Physiol Endocrinol Metab. 2024 Apr 1;326(4):E454-E471. doi: 10.1152/ajpendo.00189.2023. Epub 2023 Dec 6.
7
Likelihood-ratio test statistic for the finite-sample case in nonlinear ordinary differential equation models.非线性常微分方程模型有限样本似然比检验统计量。
PLoS Comput Biol. 2023 Sep 22;19(9):e1011417. doi: 10.1371/journal.pcbi.1011417. eCollection 2023 Sep.
8
Quantifying postprandial glucose responses using a hybrid modeling approach: Combining mechanistic and data-driven models in The Maastricht Study.使用混合建模方法定量餐后血糖反应:在马斯特里赫特研究中结合机理模型和数据驱动模型。
PLoS One. 2023 Jul 27;18(7):e0285820. doi: 10.1371/journal.pone.0285820. eCollection 2023.
9
From patterns to patients: Advances in clinical machine learning for cancer diagnosis, prognosis, and treatment.从模式到患者:癌症诊断、预后和治疗的临床机器学习进展。
Cell. 2023 Apr 13;186(8):1772-1791. doi: 10.1016/j.cell.2023.01.035. Epub 2023 Mar 10.
10
From evidence-based medicine to digital twin technology for predicting ventricular tachycardia in ischaemic cardiomyopathy.从循证医学到数字孪生技术,预测缺血性心肌病室性心动过速。
J R Soc Interface. 2022 Sep;19(194):20220317. doi: 10.1098/rsif.2022.0317. Epub 2022 Sep 21.