Event Abstract

Towards Reusable Experiments: Making Metadata While You Measure

  • 1 Carnegie Mellon University, Department of Biology, United States
  • 2 Elsevier, Elsevier Labs, Netherlands

Using research data acquired in other labs requires that the metadata, that define the conditions and manipulations of each experiment are well documented. Usually, this metadata is not stored with the experimental data itself, and is often written within the experimenter’s lab notebook and therefore is not easily searched or compiled. When scientists upload data files to a central repository (like crcns.org), metadata are often not included. For domain-specific databases (such as neuromorpho.org), metadata are added by expert curators in an expensive and painstaking process that does not scale up to the large amounts of scientific data produced every day. An effective way “scale-up” is to convince researchers to create digital metadata in real-time during their experiment. Here, we have developed an electronic lab notebook application (running on tablet computers and smartphones) to annotate in vitro electrophysiological recordings with essential methodological details.

We tailored our system to the workflows used in the collection of electrophysiology data by the Urban, Gittis, and Barth Labs at Carnegie Mellon University. These labs study a variety of brain areas, addressing hypotheses from neural coding and synaptic plasticity to the mechanisms underlying neurological disorders; however, they share a core set of methodologies associated with recording neural activity from brain slices. Using the developed app, individual experimenters enter details (like the animal strain used or the neuron type recorded) through a series of drop-down menus (see Figure 1 for a screenshot). This structured data entry approach allows us to enforce a common metadata format and the usage of INCF standards and terminologies. Additionally, we designed the app’s interface to ensure simple, efficient data entry by the user.

The collected metadata is uploaded directly to a relational database and combined with the acquired electrophysiology data files into a semantically-enriched, reusable format that allows for creative data exploration. This data can be used by the person collecting the data or others in the lab for testing hypotheses and analyzing collections of data from his or her own lab, in a web-based ‘Data-Dashboard’. Rather than being limited to datasets collected within a single lab, researchers can now find (using metadata as a search filter) and analyze relevant data collected in other labs. Through improving data organization, archiving, and sharing practices, this system will show clear benefits to the scientists performing and analyzing research data and, we hope, will empower demonstrably better neuroscience research.

Figure 1

Acknowledgements

This work was funded by the NIDCD (R01DC0005798) and Elsevier Research Data Services.

Keywords: neuron electrophysiology, digital lab notebook, data sharing, experimental metadata, Reproducible Research

Conference: Neuroinformatics 2013, Stockholm, Sweden, 27 Aug - 29 Aug, 2013.

Presentation Type: Oral presentation

Topic: General neuroinformatics

Citation: Tripathy SJ, De Waard A, Gerkin RC, Marques D, Burton SD and Urban NN (2013). Towards Reusable Experiments: Making Metadata While You Measure. Front. Neuroinform. Conference Abstract: Neuroinformatics 2013. doi: 10.3389/conf.fninf.2013.09.00051

Received: 08 Apr 2013; Published Online: 11 Jul 2013.

* Correspondence: Mr. Shreejoy J Tripathy, Carnegie Mellon University, Department of Biology, Pittsburgh, PA, 15217, United States, stripat3@gmail.com

© 2007 - 2019 Frontiers Media S.A. All Rights Reserved