FUnctionality Sharing In Open eNvironments
Heinz Nixdorf Chair for Distributed Information Systems

REPRODUCE-ME: Reproducibility of Scientific Experiments

The official website for the reproducibility of scientific experiments research: https://w3id.org/reproduceme/research

The source ontology for REPRODUCE-ME ontology is hosted on here and it is also available from the NCBO BioPortal.

The REPRODUCE-ME ontology is developed to represent the whole picture of a scientific experiment describing the path it took from its design to result. It is extended from PROV-O to represent all agents, activities and entities involved in an experiment. It also extends from P-Plan to represent the steps, the input and output variables and the complete path taken from an input to an output of an experiment. We aim to enable end-to-end reproducibility of scientific experiments by capturing and representing their provenance using the REPRODUCE-ME ontology. In order to this, we collected a list of competency questions that are required to be answered for reproducing scientific experiments from the scientists from the Collaborative Research Centers ( CRC ReceptorLight). Some of them are listed below:

  • What are the input and output variables of an experiment?
  • Which are the instruments used in an experiment and their settings when the output was generated?
  • What is the complete path taken by a scientist for an experiment?
  • Who created this experiment and when? Who modified it and when?
  • Which are the steps involved in an experiment?
  • What is the order of steps taken in an experiment?
  • Which are the scripts, publications and external resources used in an experiment?

The ontology is used to answer these kinds of competency questions.

Using the REPRODUCE-ME ontology, we can describe and semantically query the information for the following:

  • Scientific experiments
  • Input and output associated with an experiment
  • Execution Environmental attributes of an experiment
  • Experiment Materials and their preparation steps
  • Steps used in an experiment
  • Execution Order of steps and activities
  • The agents involved in an experiment and their role.
  • Script/Jupyter Notebook execution in a multi-user environment provided by JupyterHub.
  • The OME data model
  • Microscopic Image Acquisition properties
  • Bio-Formats
  • Instruments and their settings

The ontology is used and evaluated in the prototype of ReceptorLight project. The prototype is developed by extending OMERO  and provides a form-based provenance capture system so that scientists can document their experimental data along with the images, scripts, standard operating procedures, files etc.

We use ontology-based data access technique to access the data from the OMERO which is stored in PostgreSQL and semantically link to the terms in the REPRODUCE-ME ontology. The ontology-based mapping of OMERO, experimental data and Jupyter Notebook data are available in GitHub. It also provides the SPARQL queries used for the evaluation of the ontology.