SelfishGene / filter_and_fire_neuron

Code behind the work "Multiple Synaptic Contacts Combined with Dendritic Filtering Enhance Spatio-Temporal Pattern Recognition of Single Neurons", bioRxiv 2022

The Filter and Fire (F&F) Neuron Model

This repo contains the code behind the work
Multiple Synaptic Contacts combined with Dendritic Filtering enhance Spatio-Temporal Pattern Recognition capabilities of Single Neurons

Multiple Synaptic Contacts combined with Dendritic Filtering
enhance Spatio-Temporal Pattern Recognition capabilities of Single Neurons

David Beniaguev, Sapir Shapira, Idan Segev, Michael London

Abstract: A cortical neuron typically makes multiple synaptic contacts on the dendrites of a post-synaptic target neuron. The functional implications of this apparent redundancy are unclear. The dendritic location of a synaptic contact affects the time-course of the somatic post-synaptic potential (PSP) due to dendritic cable filtering. Consequently, a single pre-synaptic axonal spike results with a PSP composed of multiple temporal profiles. Here, we developed a "filter-and-fire" (F&F) neuron model that captures these features and show that the memory capacity of this neuron is threefold larger than that of a leaky integrate-and-fire (I&F) neuron, when trained to emit precisely timed output spikes for specific input patterns. Furthermore, the F&F neuron can learn to recognize spatio-temporal input patterns, e.g., MNIST digits, where the I&F model completely fails. Multiple synaptic contacts between pairs of cortical neurons are therefore an important feature rather than a bug and can serve to reduce axonal wiring requirements.

Resources

Open Access version of Paper: biorxiv.org/content/10.1101/2022.01.28.478132v2
Data required for full replication of all results: kaggle.com/selfishgene/fiter-and-fire-paper
Introductory Notebook (Figure 1 in manuscript): kaggle.com/selfishgene/f-f-introduction-figure-fig-1
Notebook with replication of main results 1: kaggle.com/selfishgene/f-f-capacity-figure-fig-2
Notebook with replication of main results 2: kaggle.com/selfishgene/f-f-mnist-figure-fig-3
Notebooks for full replication of all figures: kaggle.com/selfishgene/fiter-and-fire-paper/code

Increased capacity of F&F vs I&F

• Use create_capacity_figure_Fig2.py to replicate Figure 2 in the manuscript
  • All major parameters are documented inside the file using comments
  • All necessary files are under the folder results_data_capacity\
• Use FF_vs_IF_capacity_comparison_interactions.py to recreate all files in results_data_capacity\
  • All major parameters are documented inside the file using comments
  • Use run_capacity_configs_on_cluster_slurm.py to send jobs to a slurm cluster

Single Neurons as Spatio-Temporal Pattern Recognizers

• Use create_MNIST_figure_Fig3.py to replicate Figure 3 in the manuscript
  • All major parameters are documented inside the file using comments
  • All necessary files are under the folder results_data_mnist\. large files are on the dataset on kaggle
• Use MNIST_classification_LR_IF_FF_interactions.py to recreate all files in results_data_mnist\
  • All major parameters are documented inside the file using comments
  • Use run_mnist_configs_on_cluster_slurm.py to send jobs to a slurm cluster

PSPs of a realistic detailed biophysical Layer 5 Cortical Pyramidal Neuron

• Visit this link to replicate Supplementary Figure S2 in the manuscript
• All necessary simulation data for this figure are in the file sim_results_excitatory.p in the dataset on kaggle

Acknowledgements

We thank all lab members of the Segev and London Labs for many fruitful discussions and valuable feedback regarding this work. In particular we would like to thank Sapir Shapira that skillfully collected all data and created Supplementary Figure S2 in the paper.

If you use this code or dataset, please cite the following work:

1. David Beniaguev, Sapir Shapira, Idan Segev and Michael London. "Multiple Synaptic Contacts combined with Dendritic Filtering enhance Spatio-Temporal Pattern Recognition capabilities of Single Neurons ." bioRxiv 2022.01.28.478132; doi: https://doi.org/10.1101/2022.01.28.478132