Alyx

Alyx manages centralized experimental metadata, file storage, and automated analysis pipelines to enable sharing and processing of neural activity and behavioral data across distributed neuroscience laboratories.


Key Features:

  • Modular Architecture: Integrates components to streamline data management and handle large volumes of neural activity data recorded from mice engaged in perceptual decision-making tasks.
  • Electronic Lab Notebook (Alyx): Records experimental metadata and automatically registers and uploads recorded data files to the central repository.
  • Data Access via ONE: Provides programmatic access through the Open Neurophysiology Environment (ONE) for searching and retrieving data with integration for MATLAB and Python.
  • Automated Data Analysis Pipelines: Utilizes DataJoint to run automated pipelines that process collected data and populate a website with graphical summaries.
  • Centralized Server and Database Management: Uses a setup.py script to configure PostgreSQL databases, database connection settings, and a Python virtual environment with dependencies including Django.
  • Distinct User Roles: Separates PostgreSQL administrative users from Alyx (Django) users for database maintenance and data management operations.

Scientific Applications:

  • Large-scale collaborative neuroscience (IBL): Centralizes contribution, access, and analysis of experimental data across International Brain Laboratory sites to measure neural activity across different brain regions.
  • Behavioral and neural data collection: Supports storage and organization of data from mice performing perceptual decision-making tasks for downstream analysis.
  • Programmatic analysis and modeling: Enables data retrieval and computational modeling workflows via MATLAB and Python integration.
  • Automated summary visualization: Produces graphical summaries through DataJoint pipelines to support result inspection and monitoring.

Methodology:

Uses PostgreSQL databases with a Django backend; a setup.py script configures databases, connection settings, and a Python virtual environment with dependencies including Django; ONE provides programmatic data access; DataJoint implements automated analysis pipelines that process recorded data and populate a website with graphical summaries.

Topics

Details

Programming Languages:
JavaScript, MATLAB, Python
Added:
1/14/2020
Last Updated:
12/2/2020

Operations

Publications

Bonacchi N, Chapuis G, Churchland A, Harris KD, Hunter M, Rossant C, Sasaki M, Shen S, Steinmetz NA, Walker EY, Winter O, Wells M. Data architecture for a large-scale neuroscience collaboration. Unknown Journal. 2019. doi:10.1101/827873.

Documentation