spycon: A toolbox for benchmarking neuronal connectivity inference

Introduction

Numerous methods have been proposed that attempt to infer neuronal connectivity from parallel recorded spike trains. Here we provide the spycon framework, that aims at making a comparable framework among these methods. We provide several connectivity inference algorithms, and a template, such that costum algorithms can be intergrated into spycon. spycon then provides a unified output of the methods, that methods can be benchmarked on different datasets.

Quickstart | Install guide | Citing | Documentation

Quickstart

You can either install spycon locally. If you just want to try it out quickly, you can also start and interactive session on renku and directly start coding. Check the notebooks folder for some examples.

Here is a simple example, that shows how easy it is to do connectivity inference.

from spycon.coninf import Smoothed_CCG

times, ids = ...                           # Load your data as simple 1D numpy arrays
spycon_method = Smoothed_CCG()             # Initialize your method with default parameters
spycon_result = spycon_method.infer_connectivity(times, ids) # Run inference
spycon_result.draw_graph()                 # Draw inferred graph

Installation

Clone the repository into a directory and go into the folder. Just do the following

pip install git+https://github.com/christiando/spycon

For code development do

git clone https://github.com/christiando/spycon.git
cd spycon/
pip install -r requirements.txt
pip install -e .

Implemented algorithms

All the algorithms that are discussed in our paper are implemented in this toolbox.

Method	Abbrevation	Remarks
`CoincidenceIndex`	CI	Assess significance of coincidence index via surrogate data
`Smoothed_CCG`	sCCG	Assess significance of peaks of CCG by comparing to smoothed CCG
`directed_STTC`	dSTTC	Assess signficance of directed STTC via surrogate data
`GLMCC`	GLMCC	Assess significance of peaks of CCG via parametric GLM fit
`TE_IDTXL`	TE	Assess significance of transfer entropy via surrogate data
`GLMPP`	GLMPP	Assess significance of GLM coupling by posterior approximation
`NNEnsmble`	eANN	Predict (type of) connection based on the outcome of other methods

Disclaimer: NNEnsemble is a supervised method. In our paper, we carefully designed training data, that is approximately close to the experimental conditions. If your data varies from this, performance is expected to degrade. Furthermore, each algorithm has different parameters, and depending on your dataset the default one might not be appropriate. Some benchmarking on some surrogate datasets, that resemble the regime of your experimental data will be helpful to assess this.

Using you own data

spycon aims at keeping the bar low for using the connectivity algorithms for your data. We voluntarily choose, that it is enough to provide the data in the form of two simple numpy arrays, one containing the spike times (in seconds), and one the unit IDs. Note, that this makes the code very simple to use, but gives the user more responsibility of checking, that the data is in appropriately scaled, ordered, etc, compared to more sophisticated data formats.

Creating your own test

spycon contains some functionalities to make benchmarking easier on labeled data. If you have spiking data & information about the underlying connectivity, you can create an ConnectivityTest. On these objects, all implemented connectivity inference methods can be run, and provide you with some performance metrics, for benchmarking and allowing you to decide for the appropriate algorithm.

Contributing your own algorithm

If you want to use other methods than the one implemented look at the sci_template.py

christiando / spycon

readme