DataMaze: an open source methods and data repository for Morris water maze experiments

Abstract

A core limitation of behavioural neuroscience is the use of a given assay with very different apparatus specifications (e.g., size) and task parameters (e.g., retention intervals). Such procedural variability largely contributes to the outcome discrepancy across various studies and labs, thereby hindering research progress and the replicability of the outcomes (Crabbe et al., 1999). There is an immediate need for informative and inclusive quantitative benchmarks to design experiments, consider and analyze available data, and interpret the results accordingly, thereby reinforcing the FAIR principles. We took the first step in this direction by establishing a comprehensive database, starting with the Morris water maze (DataMaze). DataMaze contains and consolidates specifications and task parameters from thousands of papers (∼5000) and includes data along with related experimental manipulations. To reinforce this effort, we introduce easy-to-calculate indices that will help compare different studies in terms of parameters and outcomes, both prospectively and retrospectively. We also showcase the use of DataMaze for secondary data analysis. The ultimate aim of DataMaze is to transform research practices and reporting standards in behavioural neuroscience, such that researchers quantitatively utilize virtually all published information in evaluating their experiments, analyzing data, and interpreting results. DataMaze will remain a live and accessible platform that will grow retrospectively by incorporating information from additional studies and prospectively by accepting new and richer entries (e.g., individual subject-level data) from researchers.