Aim / Overveiw

This repository is dedicated to the analysis of single-cell RNAseq datasets, facilitating a comprehensive understanding of cell-cell communication through two main components:

1. Pre-Processing of Datasets: Jupyter Notebooks are provided to prepare the datasets for subsequent analysis using the community tool. The pre-processed data is essential for accurate and effective cell-cell communication analysis.
2. Method Comparison: This section offers a comparative study of different tools (CellPhoneDB, NicheNet) in cell-cell communication analysis. It serves to highlight the advantages and use-cases of the community tool within the broader context of available analytical methods.

Additionally, this repository is designed in alignment with the FAIR (Findable, Accessible, Interoperable, and Reusable) principles to ensure the highest standards of open science. It serves as a resource for reproducing the results, enabling researchers to validate and build upon our findings in the field of cell-cell communication

For detailed analysis using the community tool, refer to the Community Repository.

Directory Structure

Directory containing Jupyter Notebooks and the output from each section will be saved in the corresponding folder under the /outs/ directory.

Data pre-processing

We applied community tool on three published datasets.

1. Lasry et al, 2023. The publication can be accessed via the following link: https://doi.org/10.1038/s43018-022-00480-0. The raw dataset can be downloaded by running ./download_raw_data.sh --lasry.
2. Similie et al, 2019. This publication can be access via the following link: https://doi.org/10.1016/j.cell.2019.06.029. This published dataset is available for download from the controlled-access data repository, Broad DUOS, thus it requires a registration. https://portals.broadinstitute.org/single_cell/study/SCP259.
3. Integrated van Galen et al. - Oetjen et al.: To study the alterations in cell-to-cell communication in AML, we constructed an integrated dataset containing the single-cell RNA sequencing (scRNAseq) profiles of bone marrow from healthy individuals ( Oetjen et al., 2018) and AML patients at diagnosis (vanGalen et al., 2019,). The read counts and annotation files were acquired from GSE116256 and GSE120221 datasets. These both datasets can be downloaded by running ./download_raw_data.sh --vanGalen_Oetjen.

NOTE: If you want to replicate the results and skip the preprocessing step of each dataset, you can download the pre-processed datasets from the following Zenodo links.

Pre-processed data for Lasry et al, 2023

Pre-processed data for Similie et al, 2019

Pre-processed data for Integrated van Galen et al. - Oetjen et al

Method comparison

This section is organized into four main parts, each designed to evaluate the community tool in relation to other established cell-cell communcation analysis tools.

1. Comparing Default Databases: This part involves standardizing the databases to a common format to address the heterogeneity in gene space and ligand-receptor pair. Subsequently, we visualize and check the differences.
2. Building a Unified Database and Running Algorithms: We generate a customized database using community tool's database, ensuring each tool operates utilizing the identical database.
3. Comparing Results: To facilitate a direct comparison, we first transform the output from each tool into a standardized format, overcoming the challenge of different results structures.
4. Resource Usage: The robustness and scability of the tools are put to the test through subsampling across a range from 6 to 32 scRNAseq samples, measuring the tools efficiency and adaptability to varying data sizes.

How to run

• Clone the repo git clone https://github.com/colomemaria/community-paper.git and then cd into the directory cd community-paper/

• If you don't have conda installed yet, install conda by running the command below

  make install-conda
  

• Create a conda environment named "community_paper" and install all necessary packages by using the following command:

  make create-env
  

• Launch Jupyter to access the notebooks to generate graphs

  make run-jupyter
  

• Go to http://localhost:8888 (a page should open automatically in your browser)