This title appears in the Scientific Report :
2017
Towards reproducible workflows for electrophysiology data using the Elephant analysis framework
Towards reproducible workflows for electrophysiology data using the Elephant analysis framework
The degree of complexity when working with data from electrophysiological experiments has reached a level where well-structured and defined workflows for data and metadata acquisition, pre-processing, and subsequent analysis are becoming a necessity. The implementations of such workflows are often h...
Saved in:
Personal Name(s): | Denker, Michael (Corresponding author) |
---|---|
Yegenoglu, Alper / Grün, Sonja | |
Contributing Institute: |
Jara-Institut Brain structure-function relationships; INM-10 Computational and Systems Neuroscience; IAS-6 Computational and Systems Neuroscience; INM-6 |
Imprint: |
2017
|
Conference: | 12th Meeting of the German Neuroscience Society, Göttingen (Germany), 2017-03-22 - 2017-03-25 |
Document Type: |
Abstract |
Research Program: |
Supercomputing and Modelling for the Human Brain Human Brain Project Specific Grant Agreement 1 Theory, modelling and simulation |
Publikationsportal JuSER |
The degree of complexity when working with data from electrophysiological experiments has reached a level where well-structured and defined workflows for data and metadata acquisition, pre-processing, and subsequent analysis are becoming a necessity. The implementations of such workflows are often heterogeneous across researchers and experiments, dependent on custom-written codes, and far from being automatized. As such, they place a high burden and workload on the individual researchers in charge of defining the workflows and translating them into software. While in the meantime a number of generic software solutions to support some parts of such workflows are under development, software covering other aspects of the workflow are still lacking. This situation has serious consequences regarding the degree of reproducibility of data capture and data analysis in that it leads to ineffective and unsustainable science. Our aim is therefore to create and refine guidelines and tools that facilitate transparent and accessible workflows for managing and analyzing electrophysiological data.Here we outline how already today existing software tools can be combined to construct partial workflows that are capable of addressing some of the resulting challenges facing researchers, as summarized in [1]. To this end, we introduce a case study that links emerging software tools to form a reproducible analysis workflow based on the Python programming language. At the heart of this workflow we identified and partly developed three open-source software tools that represent the scaffold from which the analysis is built. First, we demonstrate how data of different origins can be represented in a standard form using the [*i*]Neo[*/i*] framework [2]. Second, we demonstrate how the complex metadata accumulating in an electrophysiological experiment [3] can be gathered and stored using the open metadata markup language ([*i*]odML[*/i*]) for metadata management [4]. These metadata are suitable to be combined with the actual data in the Neo framework, leading to a common representation of both data and metadata for subsequent use in the analysis workflow. Finally, as the key component of such workflows, we introduce the Electrophysiology Analysis Toolkit ([*i*]Elephant[*/i*], http://neuralensemble.org/elephant/) as a recent community-centered initiative to develop an analysis framework for multi-scale activity data based on this data representation. As such, [*i*]Elephant[*/i*] represents a modular software component that provides generic library functions to perform standard and advanced analysis processes. In an outlook, we outline how this workflow can be extended by additional tools and technologies to handle access to high-performance computing, provenance tracking of the results, and work in a highly collaborative environment (see also [5]). In part, the work presented in this abstract is detailed in [6].[*b*]References:[*/b*][1] Badia, R., Davison, A., Denker, M., Giesler, A., Gosh, S., Goble, C., Grewe, J., Grün, S., Hatsopoulos, N., LeFranc, Y., Muller, J., Pröpper, R., Teeters, J., Wachtler, T., Weeks, M. & Zehl, L. (2015) INCF Program on Standards for data sharing: new perspectives on workflows and data management for the analysis of electrophysiological data. Techn. Report, International Neuroinformatics Coordination Facility (INCF). https://www.incf.org/about-us/history/incf-scientific-workshops[2] Garcia, S., Guarino, D., Jaillet, F., Jennings, T., Pröpper, R., Rautenberg, P.L., Rodgers, C.C., Sobolev, A., Wachtler, T., Yger, P. & Davison, A. (2014) Neo: an object model for handling electrophysiology data in multiple formats. Frontiers in Neuroinformatics 8, 10.[3] Zehl, L., Jaillet, F., Stoewer, A., Grewe, J., Sobolev, A., Wachtler, T., Brochier, T., Riehle, A., Denker, M. & Grün, S. (2016) Handling Metadata in a Neurophysiology Laboratory. Frontiers in Neuroinformatics, 10, 26.[4] Grewe, J., Wachtler, T. & Benda, J. (2011) A Bottom-up Approach to Data Annotation in Neurophysiology. Frontiers in Neuroinformatics 5, 16.[5] Yegenoglu, A., Senk, S., Amblet, O., Brukau, Y, Davison, A., Lester, D., Lührs, A., Quaglio, P., Rostami, V., Rowley, A., Schuller, B., Stokes, A, van Albada, S.J., Zielasko, D., Diesmann, M., Weyers, B., Denker, M. & Grün, S. An Exemplary Collaborative Simulation-Analysis Workflow for Computational Neuroscience using HPC, JARA-HPC-Symposium (2016), Springer Series Lecture Notes in Computer Science (in press).[6] Denker, M. & Grün, S. Designing workflows for the reproducible Analysis of Electrophysiological Data. In: Brain Inspired Computing, eds: Katrin Amunts, Lucio Grandinetti, Thomas Lippert, Nicolai Petkov. Lecture Notes in Computer Science, Springer (in press). |