Topology of synaptic connectivity constrains neuronal stimulus representation, predicting two complementary coding strategies.

Michal Zochowski,Michael W Reimann,Henri Riihimäki,Ran Levi,Christoph Pokorny,Jānis Lazovskis,Jason P Smith

doi:10.1371/journal.pone.0261702

Abstract

In motor-related brain regions, movement intention has been successfully decoded from in-vivo spike train by isolating a lower-dimension manifold that the high-dimensional spiking activity is constrained to. The mechanism enforcing this constraint remains unclear, although it has been hypothesized to be implemented by the connectivity of the sampled neurons. We test this idea and explore the interactions between local synaptic connectivity and its ability to encode information in a lower dimensional manifold through simulations of a detailed microcircuit model with realistic sources of noise. We confirm that even in isolation such a model can encode the identity of different stimuli in a lower-dimensional space. We then demonstrate that the reliability of the encoding depends on the connectivity between the sampled neurons by specifically sampling populations whose connectivity maximizes certain topological metrics. Finally, we developed an alternative method for determining stimulus identity from the activity of neurons by combining their spike trains with their recurrent connectivity. We found that this method performs better for sampled groups of neurons that perform worse under the classical approach, predicting the possibility of two separate encoding strategies in a single microcircuit.

Highlights

Advances in experimental techniques have allowed us to record the in-vivo activity of hundreds of neurons simultaneously
Synaptic connectivity and stimulus representation as we found earlier that taking the neighborhood structure of a sample into account led to better results, we used an alternative approach: Using only the neurons in a single volumetric sample, we extracted the 25 largest neighborhoods that could be found within a single volumetric sample, referred to as sub-neighborhoods, which were analyzed as the other neighborhood samples (Fig 7D2)
We have confirmed that the emergent activity of the NMC-model can be described according to the manifold hypothesis, that is, as determined by a lower-dimensional space; and that the values of those dimensions encode information about inputs given to the model

Summary

Introduction

Advances in experimental techniques have allowed us to record the in-vivo activity of hundreds of neurons simultaneously. This has gone along with new processing techniques to extract information from such spike trains. Some are based on the manifold hypothesis [1], which describes the high-dimensional spiking activity of large neuron populations as determined by a lower-dimensional space whose components may be aligned with behavioral (such as movement direction) or stimulus variables. The reconstruction of the underlying space can improve the decoding of such variables.

Methods

Results

Conclusion