Error-Robust Modes of the Retinal Population Code.

Jason S Prentice,Michael J Berry,Olivier Marre,Gašper Tkačik,Mark L Ioffe,Adrianna R Loback,Ian H Stevenson

doi:10.1371/journal.pcbi.1005148

Abstract

Across the nervous system, certain population spiking patterns are observed far more frequently than others. A hypothesis about this structure is that these collective activity patterns function as population codewords–collective modes–carrying information distinct from that of any single cell. We investigate this phenomenon in recordings of ∼150 retinal ganglion cells, the retina’s output. We develop a novel statistical model that decomposes the population response into modes; it predicts the distribution of spiking activity in the ganglion cell population with high accuracy. We found that the modes represent localized features of the visual stimulus that are distinct from the features represented by single neurons. Modes form clusters of activity states that are readily discriminated from one another. When we repeated the same visual stimulus, we found that the same mode was robustly elicited. These results suggest that retinal ganglion cells’ collective signaling is endowed with a form of error-correcting code–a principle that may hold in brain areas beyond retina.

Highlights

Understanding the manner in which population of neurons encode information is fundamental to systems neuroscience
In conditions of high noise, an error-correcting organization featuring discrete clusters of population words optimizes information transfer. Is this the case in real neural populations? In the present paper, we address this question empirically by fitting a statistical model to the retinal ganglion cell population representing a small visual patch, and show that a discrete population structure emerges from the model and naturally captures qualitatively important features of the visual stimulus
Our model is highly tractable to fit, and allows for exact sampling from the probability distribution over retinal population response patterns. We found that this model provided an excellent description of the frequency of occurrence of activity patterns in the neural population, and we further investigated its implications for information encoding

Summary

Introduction

Understanding the manner in which population of neurons encode information is fundamental to systems neuroscience. We address this question empirically by fitting a statistical model to the retinal ganglion cell population representing a small visual patch, and show that a discrete population structure emerges from the model and naturally captures qualitatively important features of the visual stimulus. We investigated a model in which the complex, fluctuating patterns of spikes across the population could be grouped into a modest number of “collective modes.” The possibility of this dimensionality reduction is a consequence of redundancy in the retinal code. Our model is highly tractable to fit, and allows for exact sampling from the probability distribution over retinal population response patterns We found that this model provided an excellent description of the frequency of occurrence of activity patterns in the neural population, and we further investigated its implications for information encoding. We note that nothing in our analysis methods was specific to the retina, and so this approach can be readily applied to population codes in other brain regions as well

Results

Discussion

Ethics statement

Experimental procedures

Y mode sequences T

X pixels time