Single Cell Explorer (ver 1.1.0)

Thank you for visiting wiki of Single Cell Explorer. Singel Cell Explorer is available open-source under the GNU LGPLv3 license. This web application only run on Linux.

Authors: Di Feng, Dechao Shan

Contact: [email protected]

Site URL:

http://singlecellexplorer.org

http:18.204.165.197

Source code, database files, python scripts for Jupyter notebook

http://18.204.165.197/download.html

Jupyter Notebook walkthrough

http://18.204.165.197/analysis.html

Installation guide

http://18.204.165.197/install.html

Installation of Single Cell Explorer from Source

quickest way: run setup shell scripts from http://18.204.165.197/downloads/setupSCexplorer.sh

From console (step by step)

Update (based on ubuntu 18.04 or 16.04)

sudo apt-get update

sudo apt-get install -y build-essential

sudo apt-get install -y ssh libssl-dev libffi-dev libxml2-dev libxslt1-dev zlib1g-dev zip unzip libfftw3-dev libcurl3 openssl

Install python3

sudo apt-get install -y python3 python3-pip python3-dev

Install MongoDB, create database paths, and start MongoDB server

cd ~

mkdir -p mongodb

cd mongodb

wget https://fastdl.mongodb.org/linux/mongodb-linux-x86_64-ubuntu1604-4.0.10.tgz

tar -zxvf mongodb-linux-x86_64-ubuntu1604-4.0.10.tgz

mkdir scdb

mkdir log

sudo ./mongodb-linux-x86_64-ubuntu1604-4.0.10/bin/mongod --dbpath "./scdb" --port 27017 --wiredTigerCacheSizeGB 1 --fork --logpath "./log/scdb.log"

Load database schema & sample data into MongoDB

cd ~/mongodb/

wget http://18.204.165.197/downloads/scDB.zip

unzip scDB.zip

mkdir dumpfiles

mv scDB dumpfiles

mongodb-linux-x86_64-ubuntu1604-4.0.10/bin/mongorestore dumpfiles

Launch Single Cell Explorer web application

sudo pip3 install --upgrade pip

sudo pip3 install numpy gunicorn pymongo sklearn pandas django==2.2 torchvision

cd ~

mkdir singleCell

cd singleCell

wget http://18.204.165.197/downloads/singleCellExplorer.zip

unzip singleCellExplorer.zip

cd singleCellExplorer

python3 manage.py runserver 0.0.0.0:8000

usage

--For Data registration: The following items are mandatory. study: The name of the study, which should include all the samples as a collection. species: Human, Mouse, etc tissue: The biological source of the samples (blood, inflamed, uninvolved etc). mapType: Currently, we support tsne, umap, and phate. name: The map name will be used in the single cell explorer map viewer.

The following information is optional, but we encourage you to use. disease: This help to create a collected atlas of normal tissue or disease tissue. source: You can use internal or external to distinguish the data source. author: This indicate the contact person of the data or author who created the map in single cell explorer. subjectid: Subject ID, which represent the each donor in the study, should be unique. Sometime, there are multiple tissue samples could be collected from the same individual or subject ID.

Usage: For a study that collect blood, uninvolved, and involved samples from multiple subjects. We can create the following dictionary. You can also add more meta information in the dictionary.

mapinfo={
“study”:”Disease Collection”, “species”: “Homo sapien”, “tissue”: “involved”, “mapType”:”tsne”, “name”:”involved sample”, “source”:”public data”, “author”:”Me”, “subjectid”:”CT0001”, “comment”:””
}

--Multiple Sample Comparison Once you registered all the samples from a study. You can cross-compare gene expression among multiple donor/subjects within that study. You can select 1) the study 2) the right tissue to compare, 3) the cell type 4) the gene of interest for comparison.