Detection vs. discrimination in ON and OFF parasol RGCs and human observers

This repository includes the analysis code for: Human retina trades single-photon detection for high-fidelity contrast encoding

@article {Kilpelainen2022human,
  author = {Markku Kilpel{\"a}inen and Johan West{\"o} and Anton Laihi and Daisuke Takeshita and Fred Rieke and Petri Ala-Laurila},
  title = {Human retina trades single-photon detection for high-fidelity contrast encoding},
  elocation-id = {2022.12.15.520020},
  year = {2022},
  doi = {10.1101/2022.12.15.520020},
  publisher = {Cold Spring Harbor Laboratory},
  URL = {https://www.biorxiv.org/content/early/2022/12/15/2022.12.15.520020},
  eprint = {https://www.biorxiv.org/content/early/2022/12/15/2022.12.15.520020.full.pdf},
  journal = {bioRxiv}
}

Getting started:

1. Clone the repository.
2. Install the following software for full functionality:
   • MATLAB for analyses and figure panels (tested with 2017b and 2020b)
   • export_fig, required for saving publication ready panels and figures in pdf format. Install by cloning the export_fig repository to a directory under the root directory for this repository. setPaths.m (below) will then automatically add it to the MATLAB path. Figure panels are saved as png images instead of pdfs if export_fig is not found.
   • ghostscript, required for saving figures in pdf format with export_fig (gets installed automatically with LaTeX).
   • LaTeX, required for joining panels into the final manuscript figures.
   • Make, required for running all figure creating scripts automatically.
3. Download data.
   • Download data from: https://doi.org/10.6084/m9.figshare.21696686
   • Copy the following directories RGC Data, RGC Data RF, and Psychophysics Data to the root directory of this repository.
4. Run setPaths.m:
   • Resets Matlab's paths to default values so as to avoid conflicts wiht other projects.
   • Adds all folders and subfolders in this direcotry to the path.
   • Creates any missing project folders and adds then to the path.
5. Run mainRGCAnalysis.m:
   • Extracts and groups spike counts in response to flashes of various intensities.
   • Performs a 2AFC analyses in both a detection and a discrimination setting.
   • Fits a hill function to the detection curve and to all discrimination curves.
   • Determines the dipper function based on 75 % correct thresholds (the just noticeable difference).
   • Plots and stores the results.
6. Run plotRGCSummary.m to see a dipper summary of all analyzed cells (optional).
7. Run mainPsychophysicsAnalysis.m:
   • Performs a 2AFC analyses on both the detection and the discrimination tasks.
   • Fits a hill function to the detection curve and to all discrimination curves.
   • Determines the dipper function based on 75 % correct thresholds (the just noticeable difference).
   • Plots and stores the results.
8. Run plotPsychophysicsSummary.m to see a dipper summary of all subjects (optional).
9. Generate the final manuscript figures.
   • Run ./updateFigures.sh to create all Matlab panels and to compile all LaTeX files.
   • Alternatively, run Figure Panels/plotPsychophysicsSummary.m to only generate figure panels.

Structure

Figure panels: Functions for plotting figure panels in the manuscript.
Helper functions: General plotting and analysis functions.
Psychophysics analysis: Functions for analyzing psychophysics data.
RGC analyses: Functions for analyzing RGC data.
Supplement figures: Functions for supplementary figures and and panels.
Theoretical model: Scripts and functions for fitting the pooling model to data and for understanding dips in general.
Tikz: LaTex (Tikz) code for compiling each figure in the manuscript.