98%
921
2 minutes
20
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.
Download full-text PDF |
Source |
---|---|
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC12313987 | PMC |
http://dx.doi.org/10.1038/s41540-025-00570-6 | DOI Listing |
PLoS One
September 2025
Department of Molecular Biology and Genetics, Faculty of Science, Koç University, Istanbul, Türkiye.
The increasing demand for efficient recombinant insulin production necessitates the development of scalable, high-yield, and cost-effective bioprocesses. In this study, we engineered a novel mini-proinsulin (nMPI) with enhanced expression properties by shortening the C-peptide and incorporating specific residue substitutions to eliminate the need for enzymatic cleavage. To optimize its production, we applied a hybrid approach combining microscale high-throughput cultivation using the BioLector microbioreactor and statistical modeling via response surface methodology (RSM).
View Article and Find Full Text PDFBMJ Open
September 2025
Neath Port Talbot Hospital, Port Talbot, Wales, UK.
Introduction: Gestational diabetes mellitus (GDM) is common in pregnancy and is increasing in prevalence. It is associated with an increased risk of maternal and perinatal complications if not diagnosed and managed early. Most guidelines suggest making a diagnosis of GDM using an oral glucose tolerance test (OGTT) between 24 and 28 weeks of pregnancy at which stage there still is an increased risk of complications.
View Article and Find Full Text PDFEur J Pharmacol
August 2025
Center of Health Administration and Development Studies, Hubei University of Medicine, Shiyan, Hubei, 442000, China. Electronic address:
Background: Glibenclamide (Gli), a second-generation sulfonylurea, has been widely used for managing type 2 diabetes mellitus (T2DM) due to its potent glucose-lowering effect and affordability. Recently, renewed scientific interest has emerged due to its potential anti-aging effects, mediated through mitochondrial function modulation and epigenetic regulation. Despite its longstanding clinical use, Gli's comprehensive real-world safety profile remains incompletely understood.
View Article and Find Full Text PDFFront Immunol
August 2025
Institute de Recherches Cliniques de Montréal (IRCM), Montréal, QC, Canada.
In the past two decades, several tissues have been generated from the differentiation of human pluripotent stem cells (hPSCs) to model development or disease, and for use in drug testing and cell replacement therapies. A frontliner of hPSC-derived tissues used in cell replacement therapies are the pancreatic cells, which have entered multiple clinical trials since 2014 for the treatment of type 1 diabetes (T1D). Despite challenges in early trials, the detection of endogenous C-peptide in recipients was encouraging.
View Article and Find Full Text PDFNPJ Syst Biol Appl
July 2025
Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands.
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.
View Article and Find Full Text PDF