Socrates

Socrates is an R package for analyzing Single-Cell Assay for Transposase Accessible Chromatin Sequencing (scATAC-seq) data. Socrates takes as input 1. barcode meta data and 2. a cell x feature sparse matrix in triplet format (see example data in inst/extdata). The main contribution of Socrates compared to previously established methods is a regularized quasi-binomial logistic regression for single-cell chromatin accessibility for normalizing accessibility profiles across peaks and cells.

Coming updates to Socrates will provides tool for several processing and analytical steps:

1. Clustering
2. Batch effect removal
3. Cell-type annotation
4. Co-accessibility
5. Motif analysis
6. Gene accessibility
7. Pseudotime
8. scRNA-seq integration

and much more!

All users need to begin is a counts matrix (binarized) in triplet format, or a BED file (Columns 1-3: locations of Tn5 insertions, Column 4: barcode, and Column 5 strand) with genome annotation data (gff3/gtf and fai/chromosome.sizes).

---

If you use Socates in your own study, please consider citing the following article:

Alexandre P. Marand, Zongliang Chen, Andrea Gallavotti, Robert J. Schmitz. (2021). A cis-regulatory atlas in maize at single-cell resolution. Cell, doi:10.1016/j.cell.2021.04.014

---

Current release: 03/16/21 BETA v0.0.9

Installation

Socates requires R v4.0.0 or greater.

# download the devtools package if not currently installed
install.packages("devtools")
library(devtools)

# install
devtools::install_github("plantformatics/Socrates", ref="main")

Docker Image

Download the docker image

docker pull supermanwasd/rpackage_socrates

Run the docker image

docker run -p 8787:8787 -it supermanwasd/rpackage_socrates /bin/bash

Open browser and open the webpage(http://localhost:8787). The id and password both are rstudio. The image was constructed using rstudio-sever docker image (https://www.rocker-project.org/).

Thank you to YunChuan Wang for setting up the docker file.

---

Tutorials

1. Loading data, quality control, identifying cells, and creating a Socrates object

2. Loading pre-processed data, dimensionality reduction, and clustering

3. Comparison of normalization methods

4. Identify co-accessible ACRs

... more coming soon