This repository is composed of the analyses conducted in our paper,
"Neighborhood-Informed Positional Information for Precise Cell Identity Specification"

Steps:
1. Download the datasets and define the directory paths :
    a. Download the data available in the supplementary of Petkova et. al 2019 , Cell
    "Optimal Decoding Optimal Decoding of Cellular Identities in a Genetic Network"


    b. Change the directory paths to where the data is saved. Specifically, HOME_DIR.
    c. In HOME_DIR define two subdirectories, one 'Drosophila_Data' where the downloaded data she be
    and 'Droso_Res', where the decoding results will be saved to and accessed for conducting analyses
    and plotting results
2. install and import all packages included in _imports.py
3. Produce all decoding maps - probability distributions over estimated positions
given gap gene expression of local neighborhoods (neighborhood-informed) or of single positions
(cell-independent).

4. Reproduce all figures through the analyses conducted in functions included in the repository.


Files:
data directory:
    1. _data.py:
    functions pertaining to loading the expression data of the embryos.
    2. _preprocessing.py:
    functions pertaining to the preprocessing of the data such as filtering the subset of embryos
    measured in a specific time range.
    3. Data - An abstract class for handling the data and learning the likelihood distribution of gap gene expression
    given estimated position when either including neighboring expression or not.
    4. WildTypeDrosophila - Inherits from Data - implements the posterior position distribution calculation
    including learning the likelihood distribution in Wild Type Drosophila embryos.

src directory:
    1. _constants.py:
        all constant variables relevant in data loading, preprocessing, analysis, and plotting
    2. _create_datasets.py:
        function that calculate the decoding maps for both wild type and mutant data
    3. _figures.py:
        Here are the functions for plotting the panels in each figure.
    4. _imports.py:
        Includes all the package imports needed for data loading, preprocessing, analyses, and plotting
    5. _utils.py:
        includes functions for calculating different values for comparison and analyses of the decoding
    6. _visualizations.py:
       contains plot-related functions

 test_results_analysis
    1. TestResults.py
        TestResult class implementation - this data structure holds the decoding map and its relevant
        meta information.

        Functions pertaining to comparing the decoding maps such as predicting based on it, the pair rule expression
        prediction are located here.