Reducing Animal Use in Behavioral Neuroscience
Damien Huzard, PhD · Neuronautix
· 10 min
The problem
Every behavioral experiment generates control group data. In most studies, that data is published, archived, and never reused — not because no one wants to reuse it, but because the metadata that would make it queryable was never captured.
Findable: Unique identifier + machine-readable metadata in a searchable repository
Accessible: Retrievable via standard protocol — even if the data is access-controlled
Interoperable: Uses shared vocabularies and ontologies enabling integration across datasets
Reusable: Clear provenance, open license, and documentation sufficient for a second researcher without contacting the original team
Wilkinson et al., Scientific Data, 2016
Cage position within rack
Housing density and cage-change schedule
Device firmware version and calibration date
Experimenter identity (handling procedures)
Light/dark cycle timing and lux level
Ventilation gradients between rack positions confound locomotion data
Housing density changes social dynamics — unknown group size renders controls incomparable
Firmware changes alter readout algorithms
Experimenter effect is a known confound in rodent behaviour
Light intensity affects circadian rhythms and activity baseline
Neuronautix synthesis · [UNCERTAIN] — see speaker notes
"Virtual control groups use harmonized historical control data from multiple studies to replace or supplement concurrent controls — reducing animal numbers without sacrificing statistical validity."
Neuronautix synthesis from FAIR reuse principles — the principle is established; the scale of reduction depends on metadata quality and dataset availability.
Start capturing the metadata that makes tomorrow's virtual controls possible.
Design your schema before data collection begins
Link cohorts, procedures, devices, and timelines in one place
Export FAIR-ready records for repositories and VCG queries
Built for behavioral neuroscience and HCM workflows