SCIBORG - using Single-Cell data to Infer BOolean networks modeling Regulation of Genes

SCIBORG is an inferring tool of Boolean models of stages involved in a cell differentiation system using single-cell transcriptomic data

The framework searches to compute families of Boolean networks that are both compatible with scRNA-seq data and prior regulatory knowledge. It is composed of three steps:

1. The PKN reconstruction.
2. The experimental design construction.
3. The BN inference.

Prerequisites

SCIBORG only works under Python 3.8. Two submodules, included in the SCIBORG package, are required:

• pyBRAvo
• caspo

Installation

1. Clone the repository including submodules.

   git clone --recursive git@github.com:mathieubolteau/SCIBORG.git

2. Install anaconda/miniconda if not already done to create conda env. Please follow instructions here: https://docs.conda.io/en/latest/miniconda.html
3. Create a virtual env, under Python 3.8.

   cd SCIBORG
   conda env create --file environment.yml

4. Activate your virtual env.

   conda activate sciborg_env

5. Install clyngor package (an automatic solution is under development)

   pip install clyngor

6. Install SCIBORG package

   pip install .

7. Verify that SCIBORG is installed

   sciborg --help

Usage

usage: sciborg [-h] [-pkn] [-pp] [-po] [-bn] configuration_file

positional arguments:
  configuration_file    configuration file path

optional arguments:
  -h, --help            show this help message and exit
  -pkn, --pkn-construction
                        run the PKN construction
  -pp, --pseudo-perturbation-identification
                        run the pseudo-perturbation identification
  -po, --pseudo-observation-diff-maxi
                        run the pseudo-observation difference maximization
  -bn, --bn-inference   run the BN inference

SCIBORG works with a configuration file to fix the parameters of each step of the framework.

To make the PKN reconstruction step run the following command:

sciborg --pkn-construction <CONFIG FILE>

To make the Experimental design construction run the following command. Notice that step required that the previous step has already been completed.

sciborg --pseudo-perturbation-identification --pseudo-observation-diff-maxi <CONFIG_FILE>

Note that the sub-steps can be run independently.

To make the BN inference step, run the following command:

sciborg --bn-inference <CONFIG_FILE>

Demo

In the demo folder, a toy dataset to test the installation of SCIBORG and see the expected results for this toy dataset.

To reproduce the expected results, using the configuration file present in the demo/data folder, run the following steps.

1. Modify endpoint link in the demo_config.ini file. See PKN CONSTRUCTION section.

2. Run the following command:

sciborg --pkn-construction --pseudo-perturbation-identification --pseudo-observation-diff-maxi --bn-inference ./demo/data/demo_config.ini

The execution should take 3/4 minutes and produce results similar to those in demo/expected_results folder.

Author

• Mathieu Bolteau
  • @mathieubolteau
  • mathieu.bolteau.pro@gmail.com
  • Nantes Université, École Centrale Nantes, CNRS, LS2N, UMR 6004, F-44000, Nantes