Guidelines for managing scientific data have been established under the FAIR principles requiring that data be Findable, Accessible, Interoperable, and Reusable. In many scientific disciplines, especially computational biology, both data and models are key to progress. For this reason, and recognizing that such models are a very special type of 'data', we argue that computational models, especially mechanistic models prevalent in medicine, physiology and systems biology, deserve a complementary set of guidelines. We propose the CURE principles, emphasizing that models should be Credible, Understandable, Reproducible, and Extensible. We delve into each principle, discussing verification, validation, and uncertainty quantification for model credibility; the clarity of model descriptions and annotations for understandability; adherence to standards and open science practices for reproducibility; and the use of open standards and modular code for extensibility and reuse. We outline recommended and baseline requirements for each aspect of CURE, aiming to enhance the impact and trustworthiness of computational models, particularly in biomedical applications where credibility is paramount. Our perspective underscores the need for a more disciplined approach to modeling, aligning with emerging trends such as Digital Twins and emphasizing the importance of data and modeling standards for interoperability and reuse. Finally, we emphasize that given the non-trivial effort required to implement the guidelines, the community moves to automate as many of the guidelines as possible.

The SPARC Phase-2 REVA project has produced high-resolution multimodal imaging and segmentation datasets of vagus nerves from a large cohort of human subjects. All resulting datasets are in the process of being publicly released through the SPARC Portal in compliance with FAIR Data Principles. This study registers data from these published datasets into anatomical scaffolds defining a standardized vagus trunk coordinate system with per-subject radiating branch patterns, which are highly variable. This transforms the data to be independent of subject scale and normalized to anatomical landmarks, required for practical population-level and comparative studies of vagus nerve anatomy. We present techniques and results of the vagus scaffold mapping work, including overviews of the geometric normalization process, plus visualizations and quantitative information about the number, position and orientations of branches innervating selected target organs, across the whole cohort of subjects. Subject-specific vagus scaffolds are derived from ex-vivo specimens that are imaged and digitized in cut segments then digitally re-stitched. Thus, they do not capture the in-body anatomical configuration. We present a process for remapping subject-specific vagus scaffolds into the body and show visualizations of the results. This enables examination of spatial relationships between the vagus and target organs registered in the body. Finally, we describe how these FAIR datasets, together with free and open-source software tools, support the registration of fascicular and immunohistochemical data into the scaffold framework. This integrated, multiscale representation of vagus nerve anatomy provides a foundation for advancing current understanding and informing the development of neuromodulation therapies, particularly the design and placement of nerve stimulation devices. This work is supported by the NIH SPARC program under award numbers OT3OD025347 and 75N98022C00019. This abstract was presented at the American Physiology Summit 2026 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.

Cellular physiology operates at the theoretical limits set by physical laws – for example, ion channels are sensitive to the passage of a single elementary charge, and retinae can detect one photon. Eukaryotic cells also exploit every form of energy storage mechanism available – biochemical (e.g., solute concentrations and chemical bonds), electrical (e.g., capacitive charge storage in the cell membrane), mechanical (e.g., the elastic compliance of cellular membranes), and thermal (the heat storage essential for maintaining body temperature). Since cells appear to have evolved to maximally exploit the laws of physics within their particular environmental niche, we should try to explain cell behaviour in the light of those laws. This talk will present an analysis of cell homeostasis using a ‘bond graph’ modelling approach that ensures that the conservation laws of physics (i.e., conservation of mass, charge, and energy, respectively) are satisfied for the interdependent biochemical, electrical, mechanical, and thermal energy storage mechanisms operating within the cell. The bond graph approach is applied to several cell membrane transport mechanisms and then used to consider how physics constrains intracellular electrolyte homeostasis for enterocytes (the epithelial absorptive cells in the lining of the intestinal mucosa). The model includes the electrogenic sodium-potassium ATPase pump (NKA) and an inwardly rectifying potassium channel (Kir) in the basolateral membrane, the electrogenic sodium-driven glucose transporter (SGLT1) in the apical membrane, and the glucose transporter (GLUT2) expressed in both membranes.

We have developed a biophysical modelling framework in the 12 Labours project, using compartmental bond graphs and spatial scaffolds. Bond graphs allow us to represent the physiological function of proteins and functional cell units based on conservation of mass, charge and energy. Cell models of epithelial cells, cardiomyocytes, and other cell types, are used to understand how cells maintain homeostasis. To understand homeostasis at the whole-body level, the cell models are being incorporated into models of the cardiorespiratory system and the autonomic nervous system. Spatial finite element scaffolds are used to represent organ/tissue structure and whole-body anatomy. The talk will show how this physics-based multiscale modelling framework is being developed to help interpret physiological and anatomical data for the human digital twin.

Interest in the concept of a virtual human model that can encompass human physiology and anatomy on a biophysical (mechanistic) basis, and can assist with the clinical diagnosis and treatment of disease, appears to be growing rapidly around the globe. When such models are personalised and coupled with continual diagnostic measurements, they are called 'digital twins'. We argue here that the most useful form of virtual human model will be one that is constrained by the laws of physics, contains a comprehensive knowledge graph of all human physiology and anatomy, is multiscale in the sense of linking systems physiology down to protein function, and can to some extent be personalized and linked directly with clinical records. We discuss current progress from the IUPS Physiome Project and the requirements for a framework to achieve such a model.

Cardiac sympathetic innervation plays a vital role in regulating various heart functions, and its remodeling has been linked to a range of cardiovascular diseases. To understand cardiac sympathetic innervation and remodeling, it is important to elucidate the topographical distribution and morphology of sympathetic axons in the whole left ventricle. However, immunohistochemistry (IHC) on whole heart tissue presents challenges due to its thickness, as conventional methods label only small sections, disrupting tissue integrity and neural connections, and thus complicating the comprehensive analysis of sympathetic innervation. To overcome this, we previously developed a whole flat-mount technique for processing the mouse heart (ventricle thickness: 0.5-1 mm). Building on this success, we have adapted this technique to the rat heart (ventricle thickness: 3 mm). The left ventricles of Sprague-Dawley rats (adult male, n=6) were removed and prepared as flat-mounts. These samples were then processed with immunohistochemical labeling of Tyrosine Hydroxylase (TH, a sympathetic marker). Flat mounts of whole left ventricles were imaged using a Zeiss M2 Imager equipped with a 20X oil immersion lens, producing montages composed of several hundred all-in-focus projection images. For high-resolution imaging of regions of interest, a Leica confocal microscope with a 63X oil immersion lens was used. We found that: 1) TH-IR (immunoreactive) axons entered as several large bundles from the base of the left ventricles. The TH-IR axon network densely covered the whole ventricle from the base towards the apex. Specifically, individual varicose axons and terminals were found in the epicardium and myocardium, and ran along blood vessels. 2) To establish a 3D anatomical map of innervation, TH-IR axons in flat-mounts was traced, digitized, and integrated into a 3D heart scaffold using the novel Neurolucida 360 software and the organ mapping tool provided by MBF Bioscience and the NIH SPARC Map-Core. For the first time, we showed the distribution of TH-IR sympathetic axons in the whole left ventricle of rats at single cell/axon/varicosity resolution. Our results enabled us to integrate the sympathetic innervation of the entire rat ventricle onto a 3D heart scaffold at an unprecedented resolution to visualize precise anatomical distribution. This 3D model will provide the foundation for functional mapping and for understanding the remodeling of sympathetic innervation of the heart in pathological conditions, such as advanced heart failure. Supported by NIH HEAL/SPARC U01 NS113867-01 (ZJC), NIH SPARC OT3OD025349-01 (PH) and NIH 2R01HL137832 (YL Li). This abstract was presented at the American Physiology Summit 2025 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.

Chronic pain affects 25 million Americans, with 2 million developing opioid addiction due to inadequate treatment options. Visceral chronic pain, particularly abdominal pain, is a significant clinical issue; however, the neural mechanisms driving this pain are poorly understood. Our research focuses on the spinal afferent innervation of the stomach, a key pathway for visceral pain transmission. Although many studies have explored vagal afferents, the role of spinal afferents remains under-investigated, primarily due to limitations in visualization and tracing techniques. Previously, our lab determined and characterized the spinal afferent innervation in the stomach muscular wall, identifying muscle, ganglionic, and mixed-type endings. Building on this work, we conducted a study to characterize spinal afferent nerves in the submucosa of the Sprague Dawley rat stomach (male, 3-5 months). Using anterograde tracing techniques, we injected dextran-biotin into the left dorsal root ganglia (T7-T11) to label spinal afferent axons and terminals in the muscle and submucosa of the stomach. The stomach muscle and submucosa were separated from the mucosal layer, and they were further separated to produce two flat mounts of muscular and mucosal layers. Tracer-labeled spinal axons were traced using the Neurolucida tracing software, leading to detailed distribution maps of spinal afferent innervation. High-resolution photomicrographs of spinal afferent axon morphology were captured using a Zeiss M2 imager. The results were: 1) Spinal afferents were found in all regions of the stomach muscle and submucosa (corpus, fundus, antrum, cardia). 2) In the muscular layer, spinal afferent axons target specific structures, with varicose contacts with longitudinal and circular muscles, myenteric ganglia, and blood vessels. 3) In the submucosal layer, spinal afferent formed varicose axons traveling in connective tissues andalong blood vessels. 4) Spinal afferent innervation of the flat-mounts of the whole muscular and submucosal layers were mapped and intergrated into a 3D stomach scaffold. These findings bridge a gap in our understanding of spinal afferent distribution and morphology in the muscle and submucosa and provide an anatomical foundation for future functional studies and pathological remodeling. Supported by NIH 1 U01 HEAL/SPARC NS113867-01(ZC), NIH R15HL137143-01A1(ZC). This abstract was presented at the American Physiology Summit 2025 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.

Bond graphs provide a useful level of abstraction for modelling protein function for a wide range of physiological processes, such as metabolic reactions, membrane transporters, ion channels, myofilament mechanics, receptors and signalling, etc. In this talk I will show how bond graph templates (that conserve mass, charge and energy, respectively) are developed for enzyme-catalysed metabolic reactions, the sodium-potassium ATPase pump, and some of the members of the SLC transporter superfamily (including glucose transport, glutamate transport, sodium-calcium exchange, and sodium-hydrogen exchange). These can be parameterised for a specific cell and tissue type for which the experimental kinetic data is available. I will also show how analytic expressions can be derived for a representative four- or six-state model, given reasonable assumptions associated with steady state flux conditions, while always preserving thermodynamic consistency. Finally, based on work with Weiwei Ai and David Nickerson (Andre), I will present details on fitting parameters for the glucose transporters to experimental data and show how well the steady state flux expressions match the full kinetic analysis.

The SLC (solute carrier) superfamily mediates the passive transport of small molecules across apical and basolateral cell membranes in nearly all tissues. In this paper, we employ bond-graph approaches to develop models of SLC transporters that conserve mass, charge, and energy, respectively, and can be parameterized for a specific cell and tissue type for which the experimental kinetic data are available. We show how analytic expressions that preserve thermodynamic consistency can be derived for a representative four- or six-state model, given reasonable assumptions associated with steady-state flux conditions. We present details on fitting parameters for SLC2A2 (a GLUT transporter) and SLC5A1 (an SGLT transporter) to experimental data and show how well the steady-state flux expressions match the full kinetic analysis. Since the bond-graph approach will not be familiar to many readers, we provide a detailed description of the approach and illustrate its application to a number of familiar biophysical processes.

Phase 2 of the Stimulating Peripheral Activity to Relieve Conditions (SPARC) program ( https://commonfund.nih.gov/sparc ) aims to build a comprehensive map of the anatomy and functional connectivity of the human vagus nerve, allowing comparisons of inter-subject variability, and helping guide experiments on vagus nerve stimulation and the development of neuromodulation devices. Following approaches used in Phase 1 of the SPARC program, segmented nerve and fascicle data (using various imaging modalities) of vagus nerves from a large number of human subjects will be registered into an Anatomical Scaffold. This is a geometric model of the vagus nerve providing a common coordinate system across subjects. Registration is achieved by fitting a subject-specific vagus scaffold to digitized subject data, using standardized annotations for trunks, branches, fixed landmarks and orientation marks in the data and for the corresponding features of the Scaffold. We report here on progress and issues in the early stages of the study. A particular complication of the vagus is inter-subject variability. While the left and right vagus trunk are quite consistent across subjects, the number and locations of branches innervating a particular organ or target is quite different from one subject to another. This dictates generating a scaffold for each subject with its own unique branching structure. Once data is available for a reasonable number of subjects these will be remapped to a common or representative vagus scaffold. At the conclusion of this study all vagus imaging, plus derived data including the scaffold and embedded fascicles, their groupings and distributions will be freely available on the SPARC Portal ( https://sparc.science/ ). Mapping to a common scaffold allows these data to be compared and visualized at equivalent locations across all subjects in the study. This work is supported by the NIH SPARC program under award number OT3OD025347. This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

Mathematical models of cardiac cellular electrophysiology have evolved significantly over the last 50 years Beginning with the initial four ODE models from Noble in 1962, models are now being developed with many tens of differential equations requiring hundreds of parameters and integrating multiple aspects of cellular physiology Such increases in complexity inevitably result in significant barriers to the use of the models by independent scientists or even application of existing models in novel scenarios Technologies are being developed which negate these barriers to some extent and provide tools to aid model developers and users in the reuse of previous models In this chapter, we review some of the common cardiac cellular electrophysiology models and place them in the context of current model description technologies in order to Illustrate the current state of the field

Multiscale computational physiology attempts to link molecular pathways, incorporated into biophysically and anatomically based cell, tissue and organ models, to organism-wide physiological responses that are relevant in a medical context. Here we provide the background for this challenging task and discuss the development of model encoding standards and model repositories, and a mathematical modelling framework based on energy conservation principles. Functional Tissue Units (FTUs) are introduced as modular units at the tissue level to bridge the gap between molecular systems biology and anatomically based whole organ physiology.

It is important to understand the topographical distribution of nociceptive axons in the heart to treat cardiac pain. While previous studies have detected nociceptive axons in the sectioned heart, their full distribution has not been fully explored, in particular, the pattern differences in male and female mice have not been determined. In this study, we performed a topographical mapping of the nociceptive afferent axons in flat-mounts of the whole heart of mice (C57BL/6J, male & female, 3-5 months, n=6 for each gender) using calcitonin gene-related peptide (CGRP) as a marker. Then, we used a confocal microscope and a Zeiss M2 Imager microscope to scan all tissues of the heart and assemble the images into complete photo montages of the heart. Furthermore, a Neurolucida 3D Digitization and Tracing System was used to trace CGRP-immunoreactive (CGRP-IR) axons and terminals in the whole heart of the male and female mice. We found that 1) CGRP-IR axons entered as large bundles near the superior/inferior vena cava, left pre-caval vein and the pulmonary veins before bifurcating into small branches, and finally formed single varicose axons and terminals which distributed throughout the tissue, including the cardiac ganglia, SA/AV nodes, auricles, and blood vessels in the right/left atria. 2) CGRP-IR axons particularly accumulated in the SA node region. 3) CGRP-IR axons entered the right/left ventricles and interventricular septum as large bundles through the base before bifurcating into small branches towards the apex. 4) CGRP-IR axons innervated the intrinsic cardiac ganglia (ICG) neurons. 5) CGRP-IR axons wrapped around the blood vessels in the atria and ventricles. 6) CGRP-IR axons presented heavily in the epicardium layer and less in the myocardium and endocardium. 7) CGRP-IR axons had a similar innervation pattern in male and female mice. 8) The representative CGRP-IR axon innervation maps will be integrated into 3D heart scaffolds (male and female). For the first time, our data provide a comprehensive topographical map for CGRP-IR axons in the hearts of male and female mice with single cell/axon/synapse resolution. Our work lays out a novel anatomical foundation for the functional mapping of nociceptive afferent axons and their pathological remodeling in cardiovascular diseases. This work will also contribute to the development of a 3D representation of a brain-heart nociceptive afferent atlas. This study was supported by NIH HEAL/SPARC U01 NS113867-01, NIH R15HL137143-01A1. NIH SPARC OT3OD025349-01 (PH). This is the full abstract presented at the American Physiology Summit 2023 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

The function of the liver depends critically on its blood supply. Numerous in silico models have been developed to study various aspects of the hepatic circulation, including not only the macro-hemodynamics at the organ level, but also the microcirculation at the lobular level. In addition, computational models of blood flow and bile flow have been used to study the transport, metabolism, and clearance of drugs in pharmacokinetic studies. These in silico models aim to provide insights into the liver organ function under both healthy and diseased states, and to assist quantitative analysis for surgical planning and postsurgery treatment. The purpose of this review is to provide an update on state-of-the-art in silico models of the hepatic circulation and transport processes. We introduce the numerical methods and the physiological background of these models. We also discuss multiscale frameworks that have been proposed for the liver, and their linkage with the large context of systems biology, systems pharmacology, and the Physiome project. This article is categorized under: Metabolic Diseases > Computational Models Metabolic Diseases > Biomedical Engineering Cardiovascular Diseases > Computational Models.

Progress on mapping the inter-organ connectivity and intra-organ neural networks of the autonomic nervous system in several mammalian species is described. This project is one part of the NIH-funded SPARC project which is aimed at providing a quantitative description of the human autonomic nervous system to assist in the development of clinical neuromodulation devices.

Nociceptive afferent axons innervate the stomach and send signals to the brain and spinal cord. Peripheral nociceptive afferents can be detected with a variety of markers [e.g., substance P (SP) and calcitonin gene-related peptide (CGRP)]. We recently examined the topographical organization and morphology of SP-immunoreactive (SP-IR) axons in the whole mouse stomach muscular layer. However, the distribution and morphological structure of CGRP-IR axons remain unclear. We used immunohistochemistry labeling and applied a combination of imaging techniques, including confocal and Zeiss Imager M2 microscopy, Neurolucida 360 tracing, and integration of axon tracing data into a 3D stomach scaffold to characterize CGRP-IR axons and terminals in the whole mouse stomach muscular layers. We found that: 1) CGRP-IR axons formed extensive terminal networks in both ventral and dorsal stomachs. 2) CGRP-IR axons densely innervated the blood vessels. 3) CGRP-IR axons ran in parallel with the longitudinal and circular muscles. Some axons ran at angles through the muscular layers. 4) They also formed varicose terminal contacts with individual myenteric ganglion neurons. 5) CGRP-IR occurred in DiI-labeled gastric-projecting neurons in the dorsal root and vagal nodose ganglia, indicating CGRP-IR axons were visceral afferent axons. 6) CGRP-IR axons did not colocalize with tyrosine hydroxylase (TH) or vesicular acetylcholine transporter (VAChT) axons in the stomach, indicating CGRP-IR axons were not visceral efferent axons. 7) CGRP-IR axons were traced and integrated into a 3D stomach scaffold. For the first time, we provided a topographical distribution map of CGRP-IR axon innervation of the whole stomach muscular layers at the cellular/axonal/varicosity scale.

Nociceptive afferent axons innervate the stomach and send signals to the brain and spinal cord. Peripheral nociceptive afferents can be detected with a variety of markers (e.g., substance P [SP] and calcitonin gene-related peptide [CGRP]). We recently examined the topographical organization and morphology of SP-immunoreactive (SP-IR) axons in the whole mouse stomach muscular layer. However, the distribution and morphological structure of CGRP-IR axons remain unclear. We used immunohistochemistry labeling and applied a combination of imaging techniques, including confocal and Zeiss Imager M2 microscopy, Neurolucida 360 tracing, and integration of axon tracing data into a 3D stomach scaffold to characterize CGRP-IR axons and terminals in the whole mouse stomach muscular layers. We found that: (1) CGRP-IR axons formed extensive terminal networks in both ventral and dorsal stomachs. (2) CGRP-IR axons densely innervated the blood vessels. (3) CGRP-IR axons ran in parallel with the longitudinal and circular muscles. Some axons ran at angles through the muscular layers. (4) They also formed varicose terminal contacts with individual myenteric ganglion neurons. (5) CGRP-IR occurred in DiI-labeled gastric-projecting neurons in the dorsal root and vagal nodose ganglia, indicating CGRP-IR axons were visceral afferent axons. (6) CGRP-IR axons did not colocalize with tyrosine hydroxylase or vesicular acetylcholine transporter axons in the stomach, indicating CGRP-IR axons were not visceral efferent axons. (7) CGRP-IR axons were traced and integrated into a 3D stomach scaffold. For the first time, we provided a topographical distribution map of CGRP-IR axon innervation of the whole stomach muscular layers at the cellular/axonal/varicosity scale.

The December 2022 release of the SPARC Portal ( https://sparc.science ) included the first significant update to the anatomical connectivity flatmaps since the portal first launched. These flatmaps provide an interactive and visual map for the display and exploration of the autonomic nervous system of Human, rat, mouse, pig, and cat ( https://sparc.science/maps ). In addition to improved anatomical ‘cartoons’ for all species and adding a female Human map, these maps, for the first time, automatically render the connectivity knowledge directly retrieved from the SPARC Connectivity Knowledge Base of the Autonomic Nervous System (SCKAN; https://sparc.science/resources/6eg3VpJbwQR4B84CjrvmyD ). SCKAN contains explicit knowledge about CNS-ANS-end organ circuitry derived from SPARC data and scientific literature, in a form that supports computational reasoning. Each flatmap consists of a manually drawn base layer for the species-specific anatomical cartoon and a layer of manually drawn tracts for the large nerves. By annotating these drawings with standard reference SPARC vocabularies consistent with SCKAN usage, software tools are then able to semantically map connectivity circuits retrieved from SCKAN to the visual representation on each species’ flatmap. Beyond the interactive visual rendering of the circuitry on the SPARC Portal, the semantic consistency between flatmaps and SCKAN powers further user interface components on the SPARC Portal to surface additional knowledge for each rendered connection. For example, comprehensive links to the literature and/or data supporting a connectivity statement can be retrieved from SCKAN and provided to Portal users. Now that tools are in place to support this automated workflow to generate the flatmaps from SCKAN knowledge, we are continuing to improve the SPARC Portal to visualise new knowledge as it becomes available. NIH Common Fund, NIH Office of the Director, Awards OT3OD025349 and OT2 OD030541 This is the full abstract presented at the American Physiology Summit 2023 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

There is much interest from the bioengineering healthcare community worldwide in the concept of a whole-body "digital twin" computational model that can be personalised and linked with data from clinical imaging and hospital-based functional measurements, as well as from a variety of wearable, implantable or home-based devices. To be effective this is going to require a new much more comprehensive and integrative approach to computational physiology. Models of subcellular biology that can take advantage of tissue biomarkers (blood, urine, etc) must be linked with measurements of genotype and included in models of higher-level physiological function at the tissue and organ scale. Surrogate (via machine learning) organ models must be included in organ systems and integrated into whole-body models that include autonomic neural and endocrine control. This talk will discuss the mathematical framework for an algorithmic energy-based approach to multiscale computational physiology that lends itself to crowd-sourcing the international effort needed to tackle the demanding requirements of a "digital" or "virtual" twin for healthcare.