Abstract
Correlations in neuronal spike times are thought to be key to processing in many neural systems. Many measures have been proposed to summarise these correlations and of these the correlation index is widely used and is the standard in studies of spontaneous retinal activity. We show that this measure has two undesirable properties: it is unbounded above and confounded by firing rate. We list properties needed for a measure to fairly quantify and compare correlations. We propose a novel measure of correlation — the tiling coefficient. This coefficient, the correlation index and 33 other measures of correlation of spike times are blindly tested for the required properties on synthetic and experimental data. On the basis of this, we propose a measure to replace the correlation index. To demonstrate the benefits of this measure, we re-analyse data from six key studies investigating the role of spontaneous retinal activity on map formation which used the correlation index to draw their conclusions. We re-analyse data from β2KO and β2(TG) mutants, mutants lacking connexin isoforms and also the age-dependent changes in wild type correlations. Re-analysis of the data using this new measure can significantly change the conclusions. It leads to better quantification of correlations and therefore better inference from the data. We hope that this new measure will have wide applications, and will help clarify the role of activity in retinotopic map formation.
List of abbreviations and notation used
- ISI
- Inter-spike interval
- IQR
- Inter-quartile range
- MEA
- multielectrode array
- T
- Recording time
- Δt
- Time-window of synchrony
- a (b)
- Vector of spike times of neuron A (or B)
- NA (NB)
- Number of spikes from neuron A (or B) in recording
- NA,B[−Δt,Δt]
- Number of spike pairs where a spike from A occurs within Δt of a spike from B
- λA (λB)
- firing rate of spike train A (or B)
- λS
- firing rate of spikes shared between two trains
- d
- bin width
- A (B)
- vector of binned spike counts of A (or B)
- W
- sliding window width
- Ā (B̄)
- Global average of A (or B)
- Ã (B̃)
- Local average of A (or B)
- A (B)
- spike train of neuron A (or B) represented as a signal
- F
- convolution kernel
- A′ (B′)
- convolution of A (or B) with F