We used an optimal family of orthogonal tapers (slepian functions). These are parameterized by their time length T and frequency bandwidth W. For chosen T and W, maximally k = 2TW−1 tapers centered in frequency are appropriate for spectral estimation. Power spectra were

estimated over 0.4 s windows centered SB431542 in vivo on deflections (Figure 5D) and correct trials of 2.5 s (Figure 6B) with time-bandwidth product TW = 2 and k = 3 tapers. The same parameters were used for measuring spike-to-spike coherence during baseline and epochs with the largest number of deflections. To enhance readability of the LFP power at high frequencies, which are masked by the 1/fn power-law decay, we normalized the power by the frequency. We thank E. Antzoulatos, M. Bosch, S. Brincat, T. Buschman, J. Cromer, C. Diogo, M. Moazami, J. Rose, J. Roy, M. Silver, and M. Wicherski for valuable discussions on the manuscript. We also thank B. Gray, K.

MacCully, M. Noble, and D. Ouellette for technical assistance and R. Marini for surgical assistance and veterinary care. This work was supported by CELEST, a National Science Foundation Science of Learning Center (NSF OMA-0835976), NIH-NINDS R01-NS035145, and RG7204 clinical trial the Human Frontiers Science Program Organization (to M.V.P). M.V.P conceived of and designed the experiment. M.V.P performed (and E.K.M supervised) training, electrophysiological recording, and data analysis. M.V.P and E.K.M wrote the paper. “
“Ca2+ GBA3 enters a cell through NMDAR channels only when presynaptic glutamate release and depolarization of the postsynaptic membrane occur simultaneously (correlated activity). Conversely, NMDAR-mediated Ca2+ influx is suppressed at voltages near the resting membrane potential (uncorrelated activity), due to Mg2+ block, a mechanism in which the pore of NMDARs is blocked by external Mg2+ ions (Mayer et al., 1984 and Nowak et al., 1984). Since Mg2+ block allows cells to discriminate between correlated synaptic inputs and uncorrelated activity, NMDARs have been proposed to function as “Hebbian coincidence detectors.” However, the behavioral

significance and molecular effects of Mg2+-block-dependent suppression of Ca2+ influx during uncorrelated activity remains unknown (Single et al., 2000). Functional NMDARs are heteromeric assemblies of an essential NR1 subunit and various NR2 subunits. Studies of NMDAR channels have demonstrated that Mg2+ block is dependent on an asparagine (N) residue at a “Mg2+ block site” located in a putative channel-forming transmembrane segment (TM2, see Figure 2A) of each subunit (Burnashev et al., 1992, Mori et al., 1992 and Single et al., 2000). Drosophila have a single NR2 homolog, dNR2, which contains a glutamine at the Mg2+ block site (Q721), and a single NR1 homolog, dNR1, which contains an N at this site (N631). A previous study has shown that the N631 residue in dNR1 is sufficient for Mg2+ block in flies ( Xia et al., 2005).