The number of tapers, determined by the amount of time and frequency smoothing, depended on the frequency range being examined. On average, for low frequencies up to 5 Hz the time window was set to fit at least 3 cycles. For the mid-range, roughly from 5 to 15 Hz, at least 5 cycles were fit within the window span. Finally, for the gamma range the time windows were adjusted to account for 10 or more full cycles. To obtain power spectra estimates, the time–frequency representations were averaged

over quasi-stationary time intervals. The coherence for a pair of LFP signals was calculated using their multitaper auto-spectral and cross-spectral estimates. The complex value of coherence click here was evaluated first based on the spectral components averaged within a 1-s

window. Next, its magnitude was extracted to produce the time-windowed estimate of the coherence amplitude. The so-called global coherence was estimated as the grand average over all pairs of LFP signals produced in the hypercolumns. The local phenomena were quantified for signals generated within the scope of the respective hypercolumn. In addition, phase locking statistics were estimated for LFPs to Afatinib concentration examine synchrony without the interference of amplitude correlations (Lachaux et al., 1999 and Palva et al., 2005). The analysis was first performed individually for theta-, alpha- and gamma-range oscillations (with 1:1 phase relation) generated during an active attractor-coding state. In addition, cross-frequency phase coupling effects were investigated in the following pairs: theta–alpha (3:1), Interleukin-3 receptor theta–gamma (9:1) and alpha–gamma

(3:1). Phase locking value n:m (PLVn:m) between two LFP signals with instantaneous phases Φx(t) and Φy(t) was evaluated within a time window of size N as PLVn:m=1N|∑i=1Nexp(j(nΦx(ti)−mΦy(ti)))|.The window length, N, was adjusted to reach the compromise between the reliability of the estimate and the stationarity of the signals under consideration – it varied between 0.5 and 1 s, and was kept constant within any comparative analysis. It should be noted that phase locking between the same frequency band components, i.e. PLV1:1, is denoted in most cases as PLV. The instantaneous phase of the signals was estimated from their analytic signal representation obtained using a Hilbert transform. Before the transform was applied the signals were narrow-band filtered with low time-domain spread finite-impulse response filters. Additionally, a nesting relationship between theta, alpha and gamma oscillations was examined by analyzing phase-amplitude coupling effects (Vanhatalo et al., 2004, Monto et al., 2008 and Penny et al., 2008). At first, LFPs were band-pass filtered in the forward and reverse directions to extract the desirable frequency components: theta (2−5 Hz), alpha (8−12 Hz) and gamma rhythms (25−35 Hz). Then, their analytic representations were extracted by applying a Hilbert transform.