TY - JOUR
T1 - Evaluation of a novel cloud-based software platform for structured experiment design and linked data analytics
AU - Juergens, Hannes
AU - Niemeijer, Matthijs
AU - Jennings-Antipov, Laura D.
AU - Mans, Robert
AU - Morel, Jack
AU - van Maris, Antonius J.A.
AU - Pronk, Jack T.
AU - Gardner, Timothy S.
PY - 2018
Y1 - 2018
N2 - Open data in science requires precise definition of experimental procedures used in data generation, but traditional practices for sharing protocols and data cannot provide the required data contextualization. Here, we explore implementation, in an academic research setting, of a novel cloud-based software system designed to address this challenge. The software supports systematic definition of experimental procedures as visual processes, acquisition and analysis of primary data, and linking of data and procedures in machine-computable form. The software was tested on a set of quantitative microbial-physiology experiments. Though time-intensive, definition of experimental procedures in the software enabled much more precise, unambiguous definitions of experiments than conventional protocols. Once defined, processes were easily reusable and composable into more complex experimental flows. Automatic coupling of process definitions to experimental data enables immediate identification of correlations between procedural details, intended and unintended experimental perturbations, and experimental outcomes. Software-based experiment descriptions could ultimately replace terse and ambiguous 'Materials and Methods' sections in scientific journals, thus promoting reproducibility and reusability of published studies.
AB - Open data in science requires precise definition of experimental procedures used in data generation, but traditional practices for sharing protocols and data cannot provide the required data contextualization. Here, we explore implementation, in an academic research setting, of a novel cloud-based software system designed to address this challenge. The software supports systematic definition of experimental procedures as visual processes, acquisition and analysis of primary data, and linking of data and procedures in machine-computable form. The software was tested on a set of quantitative microbial-physiology experiments. Though time-intensive, definition of experimental procedures in the software enabled much more precise, unambiguous definitions of experiments than conventional protocols. Once defined, processes were easily reusable and composable into more complex experimental flows. Automatic coupling of process definitions to experimental data enables immediate identification of correlations between procedural details, intended and unintended experimental perturbations, and experimental outcomes. Software-based experiment descriptions could ultimately replace terse and ambiguous 'Materials and Methods' sections in scientific journals, thus promoting reproducibility and reusability of published studies.
UR - http://resolver.tudelft.nl/uuid:3bd51393-dc45-42de-a7f6-4f664a8bdf77
UR - http://www.scopus.com/inward/record.url?scp=85054395353&partnerID=8YFLogxK
U2 - 10.1038/sdata.2018.195
DO - 10.1038/sdata.2018.195
M3 - Article
C2 - 30280721
SN - 2052-4463
VL - 5
JO - Scientific Data
JF - Scientific Data
ER -