neuro-mine 0.9.0

Identifies the contribution of behavioural and stimulus parameters to neural activity

Neuro-MINE (Model Identification of Neural Encoding) 🧠💻

Welcome to Neuro-MINE: your handy companion for processing neuronal response data!

This app allows users to use MINE as a GUI or in the command line to train a flexible, convolutional neural network (CNN) to analyze experimental datasets containing neural activity and corresponding predictors (e.g., behavioral responses).

Neuro-MINE makes an updated version of MINE (Costabile et al., 2023) available in an easy-to-use interface. This version of MINE now supports spiking data as well as episodic data. For episodic data, care is taken that model fits, Taylor decomposition, and prediction correctly handle episode boundaries. Furthermore, Neuro-MINE provides easily interpretable outputs in a model insights file that can be used as starting points for further analysis.

How to use

Read The Docs: https://neuro-mine.readthedocs.io

Acknowledgements

Authors:
Danica Matovic
Martin Haesemeyer
Jamie Costabile
Kaarthik Balakrishnan
Sina Schwinn

Publication: Costabile JD, Balakrishnan KA, Schwinn S, Haesemeyer M. Model discovery to link neural activity to behavioral tasks. Elife. 2023 Jun 6;12:e83289. doi: 10.7554/eLife.83289. PMID: 37278516; PMCID: PMC10310322. https://elifesciences.org/articles/83289

GitHub Repository of Original Publication: https://github.com/haesemeyer/mine_pub
Lab Website: https://www.thermofish.org/

All code is licensed under the MIT license. See LICENSE for details.
© Martin Haesemeyer, Jamie D Costabile, Kaarthik A Balakrishnan, and Danica Matovic 2020-2025
Questions may be directed to [e-mail]