Ronen Segev

General properties of transcriptional time series in Escherichia coli

Gene activity is described by the time series of discrete, stochastic mRNA production events. This transcriptional time series shows intermittent, bursty behavior. One consequence of this temporal intricacy is that gene expression can be tuned by varying different features of the time series. Here we quantify copy-number statistics of mRNA from 20 Escherichia coli promoters using single-molecule fluorescence in situ hybridization in order to characterize the general properties of these transcriptional time series. We find that the degree of burstiness is correlated with gene expression level but is largely independent of other parameters of gene regulation. The observed behavior can be explained by the underlying variation in the duration of bursting events. Using Shannons mutual information function, we estimate the mutual information transmitted between an outside stimulus, such as the extracellular concentration of inducer molecules, and intracellular levels of mRNA. This suggests that the outside stimulus transmits information reflected in the properties of transcriptional time series.

Sparse low-order interaction network underlies a highly correlated and learnable neural population code

Information is carried in the brain by the joint activity patterns of large groups of neurons. Understanding the structure and function of population neural codes is challenging because of the exponential number of possible activity patterns and dependencies among neurons. We report here that for groups of ~100 retinal neurons responding to natural stimuli, pairwise-based models, which were highly accurate for small networks, are no longer sufficient. We show that because of the sparse nature of the neural code, the higher-order interactions can be easily learned using a novel model and that a very sparse low-order interaction network underlies the code of large populations of neurons. Additionally, we show that the interaction network is organized in a hierarchical and modular manner, which hints at scalability. Our results suggest that learnability may be a key feature of the neural code.

The Architecture of Functional Interaction Networks in the Retina

Sensory information is represented in the brain by the joint activity of large groups of neurons. Recent studies have shown that, although the number of possible activity patterns and underlying interactions is exponentially large, pairwise-based models give a surprisingly accurate description of neural population activity patterns. We explored the architecture of maximum entropy models of the functional interaction networks underlying the response of large populations of retinal ganglion cells, in adult tiger salamander retina, responding to natural and artificial stimuli. We found that we can further simplify these pairwise models by neglecting weak interaction terms or by relying on a small set of interaction strengths. Comparing network interactions under different visual stimuli, we show the existence of local network motifs in the interaction map of the retina. Our results demonstrate that the underlying interaction map of the retina is sparse and dominated by local overlapping interaction modules.

Synergy from Silence in a Combinatorial Neural Code

The manner in which groups of neurons represent events in the external world is a central question in neuroscience. Estimation of the information encoded by small groups of neurons has shown that in many neural systems, cells carry mildly redundant information. These measures average over all the activity patterns of a neural population. Here, we analyze the population code of the salamander and guinea pig retinas by quantifying the information conveyed by specific multicell activity patterns. Synchronous spikes, even though they are relatively rare and highly informative, convey less information than the sum of either spike alone, making them redundant coding symbols. Instead, patterns of spiking in one cell and silence in others, which are relatively common and often overlooked as special coding symbols, were found to be mostly synergistic. Our results reflect that the mild average redundancy between ganglion cells that was previously reported is actually the result of redundant and synergistic multicell patterns, whose contributions partially cancel each other when taking the average over all patterns. We further show that similar coding properties emerge in a generic model of neural responses, suggesting that this form of combinatorial coding, in which specific compound patterns carry synergistic or redundant information, may exist in other neural circuits.

Stimulus-dependent Maximum Entropy Models of Neural Population Codes

Neural populations encode information about their stimulus in a collective fashion, by joint activity patterns of spiking and silence. A full account of this mapping from stimulus to neural activity is given by the conditional probability distribution over neural codewords given the sensory input. For large populations, direct sampling of these distributions is impossible, and so we must rely on constructing appropriate models. We show here that in a population of 100 retinal ganglion cells in the salamander retina responding to temporal white-noise stimuli, dependencies between cells play an important encoding role. We introduce the stimulus-dependent maximum entropy (SDME) model—a minimal extension of the canonical linear-nonlinear model of a single neuron, to a pairwise-coupled neural population. We find that the SDME model gives a more accurate account of single cell responses and in particular significantly outperforms uncoupled models in reproducing the distributions of population codewords emitted in response to a stimulus. We show how the SDME model, in conjunction with static maximum entropy models of population vocabulary, can be used to estimate information-theoretic quantities like average surprise and information transmission in a neural population.

Inhibition of return in the archer fish

Inhibition of return is the inhibitory tagging of recently attended locations or objects. It was previously suggested that inhibition of return is a foraging facilitator in visual search. Inhibition of return was first discovered in humans and was demonstrated also in monkeys, yet it has never been demonstrated in non-primates. Here we report the presence of inhibition of return in the archer fish, which shoots down prey on overhanging vegetation, using squirts of water spouted from its mouth. Moreover, we find similar attentional effects for fish as for human participants. Our results show that the generation of inhibition of return does not require a fully developed cortex and strengthen the view that inhibition of return functions as a foraging facilitator.

Orientation saliency without visual cortex and target selection in archer fish

Our visual attention is attracted by salient stimuli in our environment and affected by primitive features such as orientation, color, and motion. Perceptual saliency due to orientation contrast has been extensively demonstrated in behavioral experiments with humans and other primates and is believed to be facilitated by the functional organization of the primary visual cortex. In behavioral experiments with the archer fish, a proficient hunter with remarkable visual abilities, we found an orientation saliency effect similar to that observed in human subjects. Given the enormous evolutionary distance between humans and archer fish, our findings suggest that orientation-based saliency constitutes a fundamental building block for efficient visual information processing.

Coding "what" and "when" in the Archer fish retina.

Traditionally, the information content of the neural response is quantified using statistics of the responses relative to stimulus onset time with the assumption that the brain uses onset time to infer stimulus identity. However, stimulus onset time must also be estimated by the brain, making the utility of such an approach questionable. How can stimulus onset be estimated from the neural responses with sufficient accuracy to ensure reliable stimulus identification? We address this question using the framework of colour coding by the archer fish retinal ganglion cell. We found that stimulus identity, “what”, can be estimated from the responses of best single cells with an accuracy comparable to that of the animals psychophysical estimation. However, to extract this information, an accurate estimation of stimulus onset is essential. We show that stimulus onset time, “when”, can be estimated using a linear-nonlinear readout mechanism that requires the response of a population of 100 cells. Thus, stimulus onset time can be estimated using a relatively simple readout. However, large nerve cell populations are required to achieve sufficient accuracy.

Pop-out in visual search of moving targets in the archer fish

Pop-out in visual search reflects the capacity of observers to rapidly detect visual targets independent of the number of distracting objects in the background. Although it may be beneficial to most animals, pop-out behaviour has been observed only in mammals, where neural correlates are found in primary visual cortex as contextually modulated neurons that encode aspects of saliency. Here we show that archer fish can also utilize this important search mechanism by exhibiting pop-out of moving targets. We explore neural correlates of this behaviour and report the presence of contextually modulated neurons in the optic tectum that may constitute the neural substrate for a saliency map. Furthermore, we find that both behaving fish and neural responses exhibit additive responses to multiple visual features. These findings suggest that similar neural computations underlie pop-out behaviour in mammals and fish, and that pop-out may be a universal search mechanism across all vertebrates.

Visual acuity in the archerfish: Behavior, anatomy, and neurophysiology

Archerfish are known for their remarkable behavior of shooting water jets at prey hanging on vegetation above water. Motivated by the fishs capacity to knock down small prey as high as two meters above water level, we studied the role of the retina in facilitating their excellent visual acuity. First, we show behaviorally that archerfish (Toxotes jaculatrix) can detect visual structures with a minimum angle of resolution in the range of 0.075°-0.15°. Then, combining eye movement measurements with a ray tracing method, we show that the image of a target on the retina coincides with the area centralis at the ventro-temporal retina. Moving down to retinal neural circuits, we then examine the ratio by which retinal ganglion cells multiplex visual information from the photoreceptors. Measuring the anatomical densities of both cell types in the area centralis, we found photoreceptor spacing to be 5.8 μm, which supports a minimum angle of resolution as low as 0.073°. Similarly, the average spacing of the ganglion cells was 5.7 μm. Based on electrophysiological measurements we found the smallest receptive fields of ganglion cells in that area to be in the range of 8-16 μm, which translates to an angular width of 0.1°-0.2°. These findings indicate that retinal ganglion cells in the area centralis stream information to the brain at a comparable resolution with which it is sampled by the photoreceptors. Thus, the archerfish can be used as an animal model for studying how visual details are streamed to the brain by retinal output.