# Computer Methods for Physiological Problems

> **NIH NIH R01** · YALE UNIVERSITY · 2024 · $406,188

## Abstract

Project Summary
 Computational neuroscience has proved itself to be a powerful companion to experimental studies; it
provides tools to assess theories of neural function and to infer dynamics across scales. This continuing
project will extend the open-source NEURON simulator, a leading platform in computational neuroscience, to
provide faster simulations, to support more types of models, and to interoperate with more tools and data
sources. The neuroscience community has used NEURON to build many new tools and neural models over its
30+ year history; over 2800 papers have reported work performed in NEURON (well over 100 in the last year).
The NEURON project seeks to increase power and ease-of-use for computational neuroscience models.
NEURON supports broad multiscale biophysical research to provide understanding of activity between, among
and within scales from cytoplasmic and extracellular molecules, through ion channels, synapses, spines,
dendrites, neurons, and local and wide-area networks. It is used directly by experimentalists and students with
little or no programming experience, providing direct mappings to familiar biological concepts through GUIs
(several alternatives). It is also used by advanced modelers who appreciate its powerful model specification
and analysis programming interface (available in Python and HOC), speed, and cross-platform support;
NEURON models run on macOS, Windows, Linux, ChromeOS, and cloud-based environments on individual
workstations and on supercomputers. User-developed components are modular and reusable, facilitating a
culture of code sharing in the community: over 800 NEURON models are available for use via the neural-
model repository ModelDB.
 We propose improvements through the following Specific Aims to provide 1: Machine-learning to reduce
multiscale, multicompartment models: massive complex models are difficult to understand, require
supercomputers to simulate, and are hard to manipulate. Our new tools will provide a principled way to obtain
reduced models from biophysical models with multiple synaptic input types and locations. Such models can be
used together with detailed models to study how changes at the cellular- or molecular-level affect large-scale
network behavior. 2: Platform for melding across simulation tools and across languages by providing
NEURON with an Application Programming Interface (API) for interoperability with other tools. The first use of
the API will be for connections with MATLAB and with musculature simulations. In future, the API will be useful
to extend NEURON use to multiple other popular languages including Julia and C++. 3: Acceleration in
numerical integration through reorganization of internals to make it possible to use GPUs and better exploit
modern hardware. This will also reduce code fragility that leads to bugs. 4: Continuing dissemination,
teaching and support with development of new documentation and tutorials, courses from undergraduate to
high technical lev...

## Key facts

- **NIH application ID:** 10936534
- **Project number:** 5R01NS011613-47
- **Recipient organization:** YALE UNIVERSITY
- **Principal Investigator:** MICHAEL L HINES
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $406,188
- **Award type:** 5
- **Project period:** 1978-07-01 → 2028-08-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10936534

## Citation

> US National Institutes of Health, RePORTER application 10936534, Computer Methods for Physiological Problems (5R01NS011613-47). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/10936534. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*