[Paper Review] Characterizing Neuronal Circuits with Spike-triggered Non-negative Matrix Factorization.

Shanshan Jia, Zhaofei Yu+4 more|arXiv (Cornell University)|Aug 12, 2018

Neural dynamics and brain function58 references3 citations

TL;DR

This study extends spike-triggered non-negative matrix factorization (STNMF) to characterize retinal neuronal circuits, revealing biophysical properties of bipolar cells—such as spatial receptive fields, temporal filters, and transfer nonlinearities—while recovering synaptic weights and separating ganglion cell spikes by presynaptic source. STNMF enables systematic decomposition of complex circuit inputs into distinct functional components.

ABSTRACT

Neuronal circuits formed in the brain are complex with intricate connection patterns. Such a complexity is also observed in the retina as a relatively simple neuronal circuit. A retinal ganglion cell receives excitatory inputs from neurons in previous layers as driving forces to fire spikes. Analytical methods are required that can decipher these components in a systematic manner. Recently a method termed spike-triggered non-negative matrix factorization (STNMF) has been proposed for this purpose. In this study, we extend the scope of the STNMF method. By using the retinal ganglion cell as a model system, we show that STNMF can detect various biophysical properties of upstream bipolar cells, including spatial receptive fields, temporal filters, and transfer nonlinearity. In addition, we recover synaptic connection strengths from the weight matrix of STNMF. Furthermore, we show that STNMF can separate spikes of a ganglion cell into a few subsets of spikes where each subset is contributed by one presynaptic bipolar cell. Taken together, these results corroborate that STNMF is a useful method for deciphering the structure of neuronal circuits.

Motivation & Objective

• To extend STNMF for systematic characterization of neuronal circuit components in retinal ganglion cells.
• To identify biophysical properties of upstream bipolar cells, including spatial receptive fields and temporal filters.
• To recover synaptic connection strengths from the STNMF weight matrix.
• To separate ganglion cell spikes into subsets based on individual presynaptic bipolar cell contributions.
• To validate STNMF as a tool for decoding complex, overlapping inputs in neuronal circuits.

Proposed method

• Apply STNMF to spike-triggered average data from retinal ganglion cells to decompose inputs into non-negative components.
• Use the non-negative factorization to extract spatial receptive fields and temporal filters of bipolar cells.
• Estimate transfer nonlinearity by analyzing the relationship between input components and spike outputs.
• Reconstruct synaptic weights from the weight matrix of the STNMF decomposition.
• Cluster spike outputs based on the dominant contributing component to identify presynaptic sources.
• Validate results by comparing recovered components with known biophysical properties of retinal circuits.

Experimental results

Research questions

• RQ1 Can STNMF accurately recover the spatial and temporal receptive fields of bipolar cells in retinal circuits?
• RQ2 How well can STNMF estimate the transfer nonlinearity of bipolar cells from spike-triggered data?
• RQ3 Can STNMF reconstruct synaptic connection strengths from neural spiking data?
• RQ4 To what extent can STNMF separate ganglion cell spikes based on individual presynaptic bipolar cell inputs?
• RQ5 Is STNMF effective in disentangling complex, overlapping inputs in a biologically plausible neuronal circuit?

Key findings

• STNMF successfully extracted spatial receptive fields and temporal filters of bipolar cells from retinal ganglion cell data.
• The method accurately estimated the transfer nonlinearity of bipolar cells, reflecting their nonlinear integration properties.
• Synaptic connection strengths were reliably recovered from the weight matrix of the STNMF decomposition.
• STNMF separated ganglion cell spikes into distinct subsets, each corresponding to a unique presynaptic bipolar cell.
• The decomposition revealed multiple functional components, confirming STNMF's ability to resolve complex input structures.
• Results demonstrate that STNMF provides a systematic and biologically interpretable method for decoding retinal circuit organization.