Event Abstract

Back to Event

Raining fire upon modelling difficulties: PyRhO in the cloud

Benjamin D. Evans1* and Konstantin Nikolic1
  • 1 Imperial College London, Institute of Biomedical Engineering, United Kingdom

Neuroscientists are accumulating data at unprecedented rates through large-scale initiatives such as the Human Brain Project (www.humanbrainproject.eu), the Allen Institute (www.alleninstitute.org) and the Japan Brain/MINDS Project (http://brainminds.jp/en/) which deposit their findings into publicly available databases. Gaining insight from this data requires modelling, analysis and sharing of results in a similarly transparent and reproducible way. However, complex installation requirements and a lack of programming experience may hinder the adoption of such modelling practices, to the detriment of our understanding of the brain. The work presented here aims to overcome these barriers to adoption through the paradigm of “Modelling as a Service” (MaaS), whereby the necessary tools are ready and waiting as a webpage in the cloud. As a proof-of-concept of the MaaS paradigm for more accessible modelling, we developed an online portal for optogenetics dubbed Prometheus (http://try.projectpyrho.org), based on PyRhO, our computational tools for fitting and simulating opsins [1]. PyRhO has already been integrated with several popular simulation environments including NEURON [2, 3] and Brian2 [4, 5], allowing simulations to be run on multiple scales (from channel to network) and includes a Graphical User Interface (GUI) based on IPython widgets. Given the surge in popularity of Python as a programming language in neuroscience, due to its expressiveness, readability, and wealth of open-source modules [6], this was a natural foundation on which to build Prometheus. More recently, the IPython/Jupyter project [7] has evolved into a browser-based notebook interface where the inputs and outputs of the modelling process are all embedded into the same webpage (along with a wide variety of other rich web content). This auto-documenting approach to simulation and analysis has been identified as an especially promising medium for promoting sharing and reproducibility in computational neuroscience [8]. As such, these technologies were natural choices for creating a browser-based modelling platform. This approach was further enhanced by the development of tmpnb which handles authentication for multiple simultaneous users, spawning separate (temporary) notebooks for each in secure Docker sandbox environments. This system is built upon a Linux server using an entirely open-source software stack to build a robust, responsive and easily replicable system, which may be rapidly deployed for both research and teaching. [Please insert Figure 1 about here] Figure 1: Screenshot of the PyRhO notebook GUI (in expanded view). The run bar is shown at the top with controls for running simulations and the step protocol tab is shown below where parameters of the stimulation protocol may be adjusted. We demonstrate Prometheus: a working web-based portal, configured with all the necessary Python dependencies preinstalled, such that they may be used on any computer with nothing more than a web browser and an internet connection. Like other online portals, this approach eliminates problems associated with local installations, such as lack of administrative permissions, version incompatibilities and unsupported operating systems to name a few. Unlike other portals, the notebook interface is lightweight, interactive and provides the opportunity to be controlled entirely through an intuitive GUI composed of familiar widgets (see Figure 1). We illustrate the advantages of such a system in terms of its interactivity, convenience and scope for easily sharing and reproducing models and analyses. The work presented greatly simplifies the use of modelling software, providing a readily accessible alternative to local installations for quick trials without the need to download, install or configure any new software. Eventually we aim to grow this project in scope to encompass other modelling and analysis tools, and migrate to Jupyter Hub for persistent individual user accounts and storage. In the meantime, we hope that this approach will serve to demonstrate how MaaS can substantially increase the appeal and accessibility of modelling, especially for scientists with a limited computational background.

Figure 1
Image

Acknowledgements

The work was kindly supported by the UK BBSRC grant: BB/L018268/1, the BBSRC Impact Acceleration Award and the UK EPSRC grant: EP/N002474/1.

References

1. Evans BD, Jarvis S, Schultz SR, Nikolic K. PyRhO: A Multiscale Optogenetics Simulation Platform. Frontiers in Neuroinformatics (2016) 10(8). doi: 10.3389/fninf.2016.00008
2. Hines M, Davison AP, Muller E. NEURON and Python. Frontiers in Neuroinformatics (2009) 3(1). doi: 10.3389/neuro.11.001.2009
3. Carnevale NT, Hines ML. The NEURON Book. Cambridge University Press (2006). doi: 10.1017/CBO9780511541612
4. Goodman DFM, Brette R. Brian: a simulator for spiking neural networks in Python. Frontiers in Neuroinformatics (2008) 2(5). doi: 10.3389/neuro.11.005.2008
5. Stimberg M, Goodman DFM, Benichoux V, Brette R. Equation-oriented specification of neural models for simulations. Frontiers in Neuroinformatics (2014) 8(6). doi: 10.3389/fninf.2014.00006
6. Muller E, Bedner JA, Diesmann M, Gewaltig M-O, Hines M, Davison AP. Python in Neuroscience. Frontiers in Neuroinformatics (2015) 9(11). doi: 10.3389/fninf.2015.00011
7. Pérez F, Granger BE. IPython: A System for Interactive Scientific Computing. Computing in Science & Engineering (2007) 9:21-29. doi: 10.1109/MCSE.2007.53
8. Topalidou M, Leblois A, Boraud T, Rougier NP. A long journey into reproducible computational neuroscience. Frontiers in Computational Neuroscience (2015) 9(30). doi: 10.3389/fncom.2015.00030

Keywords: MAAs, modelling, simulation, Portal System, Cloud computing, python, Jupyter, Prometheus, PyRhO, optogenetics, Opsin, NEURON simulator, Brian simulator, Open Science, Reproducibility of Results, Collaborative Research

Conference: Neuroinformatics 2016, Reading, United Kingdom, 3 Sep - 4 Sep, 2016.

Presentation Type: Demo

Topic: Infrastructural and portal services

Citation: Evans BD and Nikolic K (2016). Raining fire upon modelling difficulties: PyRhO in the cloud. Front. Neuroinform. Conference Abstract: Neuroinformatics 2016. doi: 10.3389/conf.fninf.2016.20.00068

Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.

The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.

Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.

For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.

Received: 31 May 2016; Published Online: 18 Jul 2016.

* Correspondence: Dr. Benjamin D Evans, Imperial College London, Institute of Biomedical Engineering, London, SW7 2AZ, United Kingdom, benjamin.evans@bristol.ac.uk