PyMICE Tools is companion to the PyMICE library for Intellicage analysis (https://github.com/Neuroinflab/PyMICE). The two are not directly affiliated. The goal of PyMICE Tools is to extend the functionality and usuability of PyMICE by providing user interfaces, a standardized method for adding high-level experimental and statistical routines, and project templates that foster reproducible research.
This is a meta repository to consolidate installation of three distinct submodules:
- PyMICE Modules (https://github.com/eturkes/pymice-modules)
- PyMICE Analyzer (https://github.com/eturkes/pymice-analyzer)
- PyMICE Server (https://github.com/eturkes/pymice-server)
PyMICE Modules is dedicated to the development of PyMICE paradigms, high-level routines as simple as calculating corner visits to analyzing following behaviors. Addition of statistical and figure generation methods, as well as miscellaneous utilities also go here. The goal is to establish a standardized API for working with the upstream PyMICE library. This has the potential of making the analysis pipeline highly modular, which will offer the user great flexibility in creating custom pipelines and implementing tools and techniques from third-party projects.
PyMICE Analyzer is a GUI and CLI project creator and pipeline manipulator. It is distinct from full-featured platforms in that it leverages the power and familiarity of standard data science tools like Jupyter and Docker. Projects are structured to suit the particular workflow of Intellicage with PyMICE and are bundled with a variety of utilities to encourage proper archivial for reproducibility. Available pipeline options are pulled directly from PyMICE Modules and included in the project structure, allowing for easy updating and shipping of specific versions with the code.
PyMICE Server has a similar workflow and functionality to PyMICE Analyzer but utilizes Wooey (https://github.com/wooey/Wooey) to create a Django-based web UI. A noteworthy advantage is that it is easily configured so that jobs are submitted to local or cloud servers for centralized, high-throughput analysis. For laboratories interesting in applying techniques with high hardware demands like machine learning, this may be a useful tool.