Events that passed 5 standard deviations (i.e., mean + 5 SD of averaged non-z-scored envelope) for more than 3 ms were considered as ripples, and ripples that were less than 20 ms apart were merged and were considered as one extended ripple. The beginning and

end of each ripple were considered as where the smoothed envelope crossed its mean value (i.e., zero for z-scored signal). Ripples events that happened when mice were not immobilized were excluded. Mice were considered as immobilized when their head speed was below 0.5 cm/s. Ripple power was obtained by applying Welch’s method on each individual non-z-scored nonenveloped ripple and then averaging over calculated powers. Morlet wavelet scalogram with bandwidth buy Cisplatin of 10 was used for spectrogram visualization of raw EEG. The same ripple-finding algorithm was also applied for gamma frequency range (25–80 Hz), to investigate whether the impairment in EEG power

is only selective for ripple events or can be found in gamma activity when animal is in immobilized state. Also, using Welch’s method, the power of raw EEG signals during run was calculated and, in particular, theta (4–12 Hz) powers for CT and KO mice were compared. For a robustness analysis, EEGs were filtered with 50-Hz-wide frequency filters ranging from 50 Hz to 600 Hz with 40 Hz overlap between two consecutive filters. http://www.selleckchem.com/products/s-gsk1349572.html Manual clustering of spikes was done based on spike waveform peak amplitude using XCLUST2 software (M.A. Wilson, MIT). Putative interneurons were also excluded from analysis on the basis of their spike width. To compare the quality of clusters in mice genotypes a modified Lratio value for each cluster of a tetrode was calculated (Pfeiffer and Foster, 2013 and Schmitzer-Torbert et al., 2005): Lratio=∑i∉C(1−CDFχdf2(Di,C2))nswhere i   ∉ C   is the set of spikes that do not belong to target cluster C

and Di,C is the Mahalanobis distance of these spikes from this cluster. CDFχdf2 is the cumulative distribution function of χ2 distribution with df = 4 (feature space for clusters is four dimensional). ns is the total number of spikes from Terminal deoxynucleotidyl transferase all the clusters (including target cluster C) of the tetrode. All the place cell analyses, except spatial coherence, were done on 1D place fields. These 1D place fields were obtained by using 2 cm bins on linear track, and these raw place fields were smoothed by applying a Gaussian smoother with a 2.4 cm SD. Place field size was calculated as the number of 2-cm-wide bins above 1 Hz threshold. Directionality index of each place field was defined as the percentage of its dominant direction (the direction that a specific cell has higher peak firing) divided by the summation of both left and right firings. Sparsity index ranges from 0 to 1, where lower value means a less diffuse and more spatially specific place field (Skaggs et al., 1996).