CX-Functional-Analysis

Code and instructions to reproduce all the analysis used to generate the figures in the article Building a functional connectome of the Drosophila central complex (Franconville, Beron, Jayaraman) and the tables underlying the associated website. All steps of the analysis are demonstrated in 3 notebooks (in the notebooks folder) that can be run here thanks to Binder.

Installation, prerequesites

To run the analysis locally (recommended if you plan of analyzing a full dataset)

• Download and install Julia. This code has been written for Julia 0.6. More recent versions will likely fail for now.
• Install the packages specified in the REQUIRE file (using Pkg.add("PkgName1","PkgName2",...))
• At the Julia prompt, type :

  Pkg.add("PlotlyJS")

• Download or clone this repository.
• Data is stored in this OpenScienceFramework page. Notebooks contain instructions to download it directly from Julia.

Content, analysis pipeline

The table LinesAndTypes.csv is identical to Table 1 of the paper and contains information about cell types and innervation patterns of the driver lines used in the study. The table labbookTable.csv is, well, a labbook (all the metadata necessary to run the analysis). Each row corresponds to one fly.

The code folder contains three scripts (mirrored in the three notebooks in the notebooks folder), all meant to be run from the CX-Functional-Analysis folder :

• functionalConnectivityRaw.jl does the movement correction, ROI detection and computes the fluorescence traces from the raw data and returns the rawData.jld file (JLD is Julia's binary format) containing a Dictionary, with one entry per fly. The script takes command line arguments : the first one is the path to where the data is located on your computer. An arbitrary number of extra arguments can be passed to specify which experimental day one want to analyze. For example julia code/functionalConnectivityRaw.jl data/full/ jun1315 jun1415 would analyze two days of experiment, assuming the data is present in the data/full/ folder.
• StatsAndExports.jl computes the statistics and summaries from the results/rawData.jld file and returns them as a series of jld (for figure making) and js (for the website) files. Notebook coming soon.
• makeFigures.jl does what you would expect. Notebook coming soon.

The workflow is schematized as :