Latest Release
Documentation
License
Build Status
Coverage
Gitter
Citation

Introduction

The Electrophys Feature Extraction Library (eFEL) allows neuroscientists to automatically extract features from time series data recorded from neurons (both in vitro and in silico). Examples are the action potential width and amplitude in voltage traces recorded during whole-cell patch clamp experiments. The user of the library provides a set of traces and selects the features to be calculated. The library will then extract the requested features and return the values to the user.

The core of the library is written in C++, and a Python wrapper is included. At the moment we provide a way to automatically compile and install the library as a Python module. Instructions on how to compile the eFEL as a standalone C++ library can be found here.

How to cite

When you use this eFEL software for your research, we ask you to cite it (this includes poster presentations) by referring to the "Cite this repository" button at the top of the repository page to get various citation formats, including APA and BibTeX.

For detailed citation information, please refer to the CITATION.cff file.

Requirements

• Python 3.9+
• Pip (installed by default in newer versions of Python)
• C++ compiler that can be used by pip
• Numpy (will be installed automatically by pip)
• The instruction below are written assuming you have access to a command shell on Linux / UNIX / MacOSX / Cygwin

Installation

The easiest way to install eFEL is to use pip

pip install efel

In case you don't have administrator access this command might fail with a permission error. In that case you could install eFEL in your home directory

pip install efel --user

Or you could use a python virtual environment

virtualenv pythonenv
. ./pythonenv/bin/activate
# If you use csh or tcsh, you should use:
# source ./pythonenv/bin/activate.csh
pip install efel

If you want to install straight from the github repository you can use

pip install git+git://github.com/BlueBrain/eFEL

Quick Start

First you need to import the module

import efel

To get a list with all the available feature names

efel.get_feature_names()

Note that the extra-cellular features, the bpap_attenuation feature and the check_ais_initiation feature are not listed above because they have to be used in a special way, as described here for extra-cellular features, here for bpap_attenuation feature and here for check_ais_initiation feature.

To change the spike detection threshold setting (default is -20 mV)

efel.set_setting('Threshold', -30)

For a full list of available settings, please refer to the Setting class

The python function to extract features is get_feature_values(...). Below is a short example on how to use this function. The code and example trace are available here

"""Basic example 1 for eFEL"""

import efel
import numpy

def main():
    """Main"""

    # Use numpy to read the trace data from the txt file
    data = numpy.loadtxt('example_trace1.txt')

    # Time is the first column
    time = data[:, 0]
    # Voltage is the second column
    voltage = data[:, 1]

    # Now we will construct the datastructure that will be passed to eFEL

    # A 'trace' is a dictionary
    trace1 = {}

    # Set the 'T' (=time) key of the trace
    trace1['T'] = time

    # Set the 'V' (=voltage) key of the trace
    trace1['V'] = voltage

    # Set the 'stim_start' (time at which a stimulus starts, in ms)
    # key of the trace
    # Warning: this need to be a list (with one element)
    trace1['stim_start'] = [700]

    # Set the 'stim_end' (time at which a stimulus end) key of the trace
    # Warning: this need to be a list (with one element)
    trace1['stim_end'] = [2700]

    # Multiple traces can be passed to the eFEL at the same time, so the
    # argument should be a list
    traces = [trace1]

    # set the threshold for spike detection to -20 mV
    efel.set_setting('Threshold', -20)

    # Now we pass 'traces' to the efel and ask it to calculate the feature
    # values
    traces_results = efel.get_feature_values(traces,
                                           ['AP_amplitude', 'voltage_base'])

    # The return value is a list of trace_results, every trace_results
    # corresponds to one trace in the 'traces' list above (in same order)
    for trace_results in traces_results:
        # trace_result is a dictionary, with as keys the requested features
        for feature_name, feature_values in trace_results.items():
            print("Feature %s has the following values: %s" %
                (feature_name, ', '.join([str(x) for x in feature_values])))


if __name__ == '__main__':
    main()

The output of this example is

Feature AP_amplitude has the following values: 72.5782441262, 46.3672552618, 41.1546679158, 39.7631750953, 36.1614653031, 37.8489295737
Feature voltage_base has the following values: -75.446665721

This means that the eFEL found 5 action potentials in the voltage trace. The amplitudes of these APs are the result of the 'AP_amplitude' feature. The voltage before the start of the stimulus is measured by 'voltage_base'. Results are in mV.

Full documentation

The full documentation can be found here

Funding

This work has been partially funded by the European Union Seventh Framework Program (FP7/2007­2013) under grant agreement no. 604102 (HBP), the European Union’s Horizon 2020 Framework Programme for Research and Innovation under the Specific Grant Agreement No. 720270, 785907 (Human Brain Project SGA1/SGA2) and by the EBRAINS research infrastructure, funded from the European Union’s Horizon 2020 Framework Programme for Research and Innovation under the Specific Grant Agreement No. 945539 (Human Brain Project SGA3). This project/research was supported by funding to the Blue Brain Project, a research center of the École polytechnique fédérale de Lausanne (EPFL), from the Swiss government’s ETH Board of the Swiss Federal Institutes of Technology.

Copyright (c) 2009-2024 Blue Brain Project/EPFL