miniML documentation

miniML documentation#

About#

This is the documentation for miniML, a machine learning framework for the analysis of synaptic events in time-series data. The documentation covers basic usage and features of miniML.

miniML uses trained deep learning models to automatically detect synaptic events in time-series data. The approach has several advantages:

  • miniML is easy to use and efficient

  • miniML offers high precision and recall

  • miniML is a general framework that can be applied to a wide range of data types and recording conditions

Usage#

miniML can be used for the following tasks:

  • Detection of synaptic events in time-series data

  • Quantitative analysis of synaptic events

  • Classification of synaptic events

  • Plotting of results

See also

The analysis method is described in detail in our manuscript, including application examples and benchmarking against previous methods.

miniML can be used for electrophysiological recordings (e.g., voltage clamp, current clamp, etc.) and other data types (e.g., calcium imaging, voltage imaging, etc.). Trained models are available for several types of data and synaptic preparations. These can be applied to a wide range of data types and recording conditions. In addition, users can train new models for their specific data.

See also

See the associated GitHub repository for source code and more information.

How to cite#

If you use miniML in your work, please cite:

@article{ONeill2024,
  title = {A deep learning framework for automated and generalized synaptic event analysis},
  url = {http://dx.doi.org/10.7554/eLife.98485.1},
  DOI = {10.7554/elife.98485.1},
  publisher = {eLife Sciences Publications,  Ltd},
  author = {O’Neill,  Philipp S. and Baccino-Calace,  Martín and Rupprecht,  Peter and Friedrich,  Rainer W. and M\"{u}ller,  Martin and Delvendahl,  Igor},
  year = {2024},
  month = jun 
}

Issues#

Please report any issues to the GitHub Issue Page.

Acknowledgements#

This documentation was built using JupyterBook.

Table of contents#

The documentation contains the following parts:

USING MINIML

GUI

MODEL TRAINING