In Trichostatin A addition, repetitive spatio-temporal patterns of firing with defined propagation schemes were identified in the network spontaneous activity. These patterns could be artificially evoked by targeted electrical and chemical stimulations. Despite the ubiquitous nature of synchronized activity patterns in neural networks, and the growing understanding of neuronal function, the manner by which a network of neurons and glia cells can give rise to synchronized activity is still under intensive research. Understanding the functional properties of neurons has evolved from a basic view of thresholddependent pulse generators that perform simple activity integration to highly complex processors that perform a variety of self regulated computational tasks. Interestingly, this evolved view of single neurons is insufficient to describe the collective dynamics and activity patterns of connected neurons. Understanding how the electrical activity properties vary upon crossover from single neurons to the network level, may provide the insight needed to reveal the innate properties of neuronal network dynamics. This understanding is particularly intriguing as it may be implemented in various fields, ranging from neural network modeling, network theory, and engineering and bio inspired devices, to name just a few. To characterize the transition from single cells to neuronal populations in terms of their electrical activity, we engineered small isolated neuro-glia clustered networks of various sizes and examined their collective activity. Clusters are of particular interest as they form spontaneously in vitro with only minimal external intervention. Moreover, clustering characterizes many biological brain networks. We promote the formation of clusters by exploiting the tendency of dissociated neuronal cells to self-organize into patterned architectures due to their preferential attachment to cell-attracting chemistries, such as poly-d-lysine or to rough surfaces, such as carbon nanotubes. By controlling their dimensions, we can systematically form and map the activity of neural networks with well identified cell numbers, ranging between several cells up to several hundreds. As we show below, such small systems demonstrate well characterized activity reflecting a clear transition from sporadic to well synchronized network level activity. Isolated small neuronal Vorinostat abmole circuits or neuronal clusters made of a few to several hundreds of neurons and glia cells were engineered using rectangular arrays of adhesive micro islands made of CNTs or PDL deposited on planar recording electrodes. Dissociated neurons and glia cells placed on such integrated multi electrode arrays or neuro-chips, self-organized into small isolated clusters with dimension between 20�C120 mm in diameter.