In our experiments this produced a higher GW4064 amplitude response in the downstream LFP at frequencies <35 Hz. At higher frequencies the amplitude of

the waveform was independent of intensity and the waveform was sinusoidal. The duty cycle and intensity of the stimulus are consequently both highly influence the waveform response, and should be carefully chosen based on the desired output. In addition, alternative temporal patterns of stimulation can also influence the neural response. Increasingly, alternative stimulation patterns are being explored for use in clinical deep brain stimulation therapies (Brocker et al., 2013). Indeed, the regimented frequency-specificity of our existing therapies and experiments appear quite artificial when compared with the natural oscillations within these neural circuits. Alternative stimulation patterns that better approximate neurologic signals, such as those presented here (Figure ​Figure77), may prove more effective in eliciting behavioral and experimental outcomes. Normal physiologic rhythms do not tend to have the frequency or phase specificity of artificial stimulation, and more varied stimuli may consequently affect neural networks differently. Poisson stimulation patterns may better reflect the stochastic firing patterns of neurons and in some cases

may prove more effective that constant-frequency stimulation (Quinkert et al., 2010; Wyckhuys et al., 2010). Cross-frequency coupling has a role in spatial memory (Shirvalkar et al., 2010), and sinusoidal stimulation could provide less synchronizing input to the neural network. The artifacts of optical stimulation that we and others have observed (Figure ​Figure88), while of significantly less magnitude than equivalent electrical stimulation artifacts, do obscure and potentially influence the underlying neurophysiologic activity. In our hands these artifacts have proven very array-dependent,

and others have suggested some mechanisms for reducing and removing them (Cardin et al., 2010). As they can prove quite insidious, leading to false detections as single units, robust methods for preventing, defining, and removing such artifacts will be necessary to limit improper conclusions. The NeuroRighter platform provides a low-cost, open-source, AV-951 real-time solution for optogenetic neuromodulation and multielectrode electrophysiology in awake and behaving animals. It is readily customizable to a number of applications, including open- and closed-loop experimentation with a variety of stimulation patterns, recording electrodes, and behavioral tasks. AUTHOR CONTRIBUTIONS Nealen G. Laxpati designed hardware adaptations, ferrules, calibration hardware and software, performed the experiments and their analysis. Jonathan P. Newman and Riley Zeller-Townson wrote the adaptations to the NeuroRighter software for open and closed-loop stimulation, and Babak Mahmoudi coded the closed-loop stimulation experiment. Claire-Anne Gutekunst and Nealen G.