Home

Analysing the electrical activity of neurons using computer code

Today in the lab, I spent a lot of my time on the computer, analysing the electrical activity of neurons that I had previously recorded during my electrophysiology experiments (see my previous post here). Today’s data analysis involved a lot of coding using Matlab software, in order to calculate various parameters such as amplitude, frequency and charge of the electrical activity of the neurons. A lot of complicated mathematical functions are needed to do these calculations, but the process is semi-automated with the use of computer code and specialised software for this type of analysis. In the end, you get some pretty graphs with the different parts of the data analysis, as shown below.

Detection of mEPSCs in a recording from a neuron. (A) Sum of two exponentials template (red) and mean event overlay (green). (B) Deconvoluted wave of the recording (blue), detection threshold (green), and detected events (red dots). (C) All-point histogram of the deconvoluted trace (blue bars) fitted with a Gaussian function (red curve). The detection threshold (green) was set to 4x the standard deviation of the Gaussian. (D) Continuous trace of detected mEPSCs (red). (E) Superimposed detected mEPSCs (grey curves) and the median mEPSC (black curve) during a continuous recording, aligned horizontally to the peak of the deconvolution function. (F) Merged events from entire recording wave with the mean ensemble (blue) and fit (red).
Detection of mEPSCs in a recording from a neuron. (A) Sum of two exponentials template (red) and mean event overlay (green). (B) Deconvoluted wave of the recording (blue), detection threshold (green), and detected events (red dots). (C) All-point histogram of the deconvoluted trace (blue bars) fitted with a Gaussian function (red curve). The detection threshold (green) was set to 4x the standard deviation of the Gaussian. (D) Continuous trace of detected mEPSCs (red). (E) Superimposed detected mEPSCs (grey curves) and the median mEPSC (black curve) during a continuous recording, aligned horizontally to the peak of the deconvolution function. (F) Merged events from entire recording wave with the mean ensemble (blue) and fit (red).

Detailed methods of today’s data analysis

Miniature excitatory postsynaptic currents (mEPSCs; the basal electrical activity of neurons) were analysed using a deconvolution-based method of detecting spontaneous mEPSCs. With this method, events are fitted with a sum of two exponentials (one for rise and one for decay) template function generated based on the time course of spontaneous synaptic events in the recorded waves. Waves are deconvolved from the template using Fourier transformation algorithms. The deconvoluted wave is fitted with a Gaussian function representing the distribution of the noise, and the detection threshold is set to four times the standard deviation of the Gaussian. The deconvoluted trace is then scanned for local maxima which correspond to the detected mEPSCs and their onset times. The median amplitude, frequency, charge, rise time and decay time were calculated for each neuron.