This repository contains the code, the notebooks, and the supplementary material for "An overview on multi-omic single-cell data joint analysis: good practices and results".
It aims to give some guidelines and good practice for whoever wants to start working with multi-Omic data analysis, specifically scRNA-seq + scATAC-seq datasets.
Martini L., Bardini R., Savino A., and Di Carlo S. An overview on multi-omic single-cell data joint analysis: good practices and results, 2023 (submitted to NAR Genomics and Bionformatics)
This Repository is uploaded on Zenodo under the DOI 10.5281/zenodo.8188656.
The Repository is alsoavailable on github at https://github.com/smilies-polito/MODA.git
This repository aims to not only share the code to obtain the results but gives step-by-step commented workflow notebooks, for the three different methods discussed in the paper.
Specifically, the repo comprises of three sections, one for each pipeline (Monocle3/Cicero, Seurat/Signac, Scanpy/Episcanpy), consisting of the notebooks with the commented code and results. For each one it is also provided a HTML image to just look at them.
One can directly access the compiled notebooks to look at the commented pipeline and results. In each pipeline subfolder one can find a html file with the corresponding pipeline workflow.
If you want to work directly on the code the Rmarkdown and jupyter files are available. They provide a cell-divided script, where each chunck can be run separately.
Depending on the language you need to set up the environment.
We made use of the renv library for ensuring reproducibility.
After installing R version 4.2.1, run the Renv (v. 0.15.5) package installation
install.packages("renv")
From inside one of the R pipeline folders, create an R project and run:
renv::init()
renv::restore()
This will take the renv.lock situated in the working directory and restore the necessary packages.
One can find the requirement file for the needed packages in the Scanpy-Episcanpy folder.
After installing them, you can work on the jupyter notebook directly.
The needed data can be retrieved by https://www.10xgenomics.com/resources/datasets/10-k-human-pbm-cs-multiome-v-1-0-chromium-controller-1-standard-2-0-0
Specifically one needs to download the following files:
- Filtered feature barcode matrix MEX (DIR)
- Filtered feature barcode matrix (HDF5)
- ATAC peak locations (BED)
- ATAC Per fragment information file (TSV.GZ)
- ATAC Per fragment information index (TSV.GZ index)
All the files need to be unzipped (except ATAC Per fragment information file (TSV.GZ)), and put on the DATA/filtered_feature_bc_matrix/ folder.
Moreover, you need to download the reference dataset for the Seurat cell-type annotation from https://atlas.fredhutch.org/data/nygc/multimodal/pbmc_multimodal.h5seurat. The file must be in the DATA/ref/ folder.
More data required are the garnett classifier (downloadable here https://cole-trapnell-lab.github.io/garnett/classifiers/) and the genomic annotations, which are provided in the DATA folder.
Since the co-accessbility calculation on the whole genome is a very long and computationally heavy calculation we provide als provide also the co-accessibility table resulted from it in the DATA/conns folder.
To run and render the markdown file run the following lines:
from the Monocle3-Cicero folder
Rscript -e "rmarkdown::render('Monocle-Cicero.Rmd')"
from the Seurat-Signac folder
Rscript -e "rmarkdown::render('Seurat-Signac.Rmd')"
Running the code will also output all the plots generated in the notebooks, saved in the Results/IMAGES folder.
|
├── Monocle3-Cicero // Monocle3 and Cicero pipeline folder
| ├── renv.lock // renv file for virtual environment setup
| ├── Monocle-Cicero.Rmd // R markdown
| └── Monocle-Cicero_html.html // HTML image
|
├── Scanpy-Episcanpy // Scanpy and Episcanpy pipeline folder
| ├── requirements.txt // requirement file for the required packages
| ├── Scanpy-Episcanpy.ipynb // Jupyter notebook
| └── Scanpy-Episcanpy.ipynb // HTML image
|
├── Seurat-Signac // Seurat and Signac pipeline folder
| ├── renv.lock // renv file for virtual environment setup
| ├── Seurat-Signac.Rmd // R markdown
| └── Seurat-Signac_html.html // HTML image
|
├── DATA // Data folder
| ├── conns // Folder for the co-accessibility matrix
| ├── filtered_feature_bc_matrix // Folder for dataset files from 10X genomic
| ├── garnett // Folder forr Garnet classifier
| ├── genomes // Folder for the genomic annotations
| └── Ref // Folder for Seurat reference dataset
|
├── Results // Results folder
| └── IMAGES // Images produced
|
├── Supplementary Material.docx // Supplementary material to the paper
|
└── README.md // This README file