Generating surrogates for significance estimation ofspatio-temporal spike patterns

Abstract

The generation of surrogate data, i.e., the modification of original data to destroy a certain feature, is used for the implementation of a null-hypothesis whenever an analytical approach is not feasible. Thus, surrogate data generation has been extensively used to assess the significance of spike correlations in parallel spike trains. In this context, one of the main challenges is to properly construct the desired null-hypothesis distribution and to avoid a bias in the null-hypothesis by altering the spike train statistics. A classical surrogate technique is uniform dithering (UD), which displaces spikes locally and uniformly. In this study, we compare UD against five surrogate techniques (two newly introduced) in the context of the detection of significant spatio-temporal spike patterns. We evaluate the surrogates for their performance, first on spike trains based on point process models with constant firing rate, and second on modeled non-stationary artificial data serving as ground truth to assess the pattern detection in a more complex and realistic setting. We determine which statistical features of the original spike trains are modified and to which extent. Moreover, we find that UD fails as an appropriate surrogate because it leads to a loss of spikes in the context of binning and clipping, and thus to a large number of false-positive patterns. The other surrogates achieve a better performance in detecting precisely timed higher-order correlations. Based on these insights, we analyze experimental data from pre-/motor cortex of macaque monkeys during a reaching-and-grasping task for spatio-temporal spike patterns. Significance statementTemporal jittering or dithering of single spikes or subsections of spike trains is a common method of generating surrogate data for the statistical analysis of temporal spike correlations. We discovered a serious problem with the classical and widely used method of uniform dithering that can lead to an overestimation of significance, i.e., to false positives in the statistical evaluation of spatio-temporal spike patterns. Therefore we consider 5 other dithering methods, compare and evaluate their statistical properties. Finally, we apply a much better method (trial shifting) to the analysis of experimental multiple-unit recordings and find several highly significant patterns that also reflect different experimental situations.