shRNAscreen

A computational pipeline for analyzing shRNA-Seq generated from in vivo shRNA screen. All 19-mers were firstly extracted from short reads of shRNA-Seq and collapsed, then searching against annotation of shRNAs sequence using R package "stringdist" with mismatches allowed.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

Prerequisites

Operating Systems

Supported Unix distributions

• Ubuntu
• CentOS
• Red Hat Enterprise Linux (please use the CentOS packages and instructions)

Job scheduler

• Univa Grid Engine
• TORQUE Resource Manager

Tools or packages

• python >= 3.5.1 (for snakemake)
• R >= 3.1.0
• pigz >= 2.3.1
• mawk >= 1.3.4
• snakemake >= 3.12.0

Libraries or modules

R

• stringdist

can be installed by following commands in R:

install.packages("stringdist");

Installing

Install snakemake

Install snakemake into a virtual environment

git clone https://bitbucket.org/snakemake/snakemake.git
cd snakemake
virtualenv -p python3 snakemake
source snakemake/bin/activate
python setup.py install

Or you can install it using bioconda

conda install snakemake

Download scripts and configuration files from github and add directory of scripts into PATH variable

git clone https://github.com/shenyang1981/shRNAscreen.git
cd shRNAscreen/; export PATH="${PWD}/scripts:$PATH"

You may consider put 'export PATH=${PWD}/scripts:$PATH' <replace ${PWD}/scripts with real path to shRNAscreen scripts> into your .bashrc file.

Prepare annotation file for shRNA library.

• hairpin_ERWOOD_good.txt -- shRNA library annotation

The annotation file includes three columns: shRNA ID, sequence and corresponding gene name (no header).

ERWOOD_1	GAGAAGATCCTCTTCATCA	Gas2
ERWOOD_2	AGCTTTGACCAGCTTCTTC	Crtc3
ERWOOD_3	GCCAGCGAATGCAGTACAT	Vmn1r181
ERWOOD_4	AGGACACATCTGCCAGCAT	Msl3
ERWOOD_5	AGCAGTACAGGCTGGTACA	Gabrr1
ERWOOD_6	AGGCTCATACTCTCCTTCT	Dcpp3
ERWOOD_7	GCCAGGATGTGACTCAGAT	Tmem171
ERWOOD_8	CATGGAGAAGTACAACATA	Dynll2
ERWOOD_9	GCCTTCATCATTGGTGCAG	Kcnj2
ERWOOD_10	ACAGAAACATTAGAATTAC	Mtpap

Prepare input files and sample information

• sampleList.txt -- information of each sequenced library, including species (Species), library ID (LibID), sequencing batch (SeqBatch), Analytic ID(AnalyticID). Samples belonged to the same AnalyticID would be selected for searching against same shRNA annotation.

The format of sampleList.txt is like:

Species	LibID	SeqBatch	AnalyticID
Mouse	AML025_rep1	AML025	batch3
Mouse	AML025_rep2	AML025	batch3

** Note: LibID should be unique as the corresponding sequence file should be named as {LibID}.fastq.gz.

• input reads files -- Reads are single-end. Name of each file should be {LibID}.fastq.gz (LibID should be the same as in sampleList.txt). All of reads files from the same sequencing batch should be put into one folder named by {SeqBatch} as indiciated in the sampleList.txt. For example, reads files "AML025_rep1.fastq.gz", "AML025_rep2.fastq.gz" can be put into folder "input/AML025/"

tree input/
input/
├── AML025
│   ├── AML025_rep1.fastq.gz
│   └── AML025_rep2.fastq.gz
└── sampleList.txt

generate config file

To generate a configuration file for snakemake, several variables need to be defined:

• SHRNASCREEN_ANADIR: path to root folder of analysis
• SAMPLEINFO: path to the sampleList.txt file
• HAIRPIN: path to the file of shRNA library annotation
• HPID: shRNA library ID of screen
• ANAID: analytic ID indicating which libraries should be selected

Configuration file can be generated using script generateConfigureFile.sh

SHRNASCREEN_ANADIR=. SAMPLEINFO=./input/sampleList.txt HAIRPIN=annotation/hairpin_ERWOOD_good.txt HPID=ERWOOD ANAID=batch3 generateConfigureFile.sh pipeline/configTemplate/conf-shRNAScreen.json > config-batch3.json

config-batch3.json

Now, files should be organized like:

.
├── annotation
│   └── hairpin_ERWOOD_good.txt
├── config-batch3.json
├── input
│   ├── AML025
│   │   ├── AML025_rep1.fastq.gz
│   │   └── AML025_rep2.fastq.gz
│   └── sampleList.txt
├── LICENSE
├── pipeline
│   ├── configTemplate
│   │   └── conf-shRNAScreen.json
│   └── counthpreads.sk
├── README.md
└── scripts
    ├── extractMatureSimple.sh
    ├── generateConfigureFile.sh
    └── summaryShCount.R

Running Pipeline

Local Mode

The pipeline can be simply run in local mode with the configuration file.

source {$pathtosnakemake}/snakemake/bin/activate
snakemake -s pipeline/counthpreads.sk --configfile config-batch3.json -j 32

Submit snakemake jobs to Cluster

Or run it by submitting to the job scheduler

Results

After running pipeline, results would be stored in "result/" and "hpreads" folder.

• AML025_rep1.mature.count.txt.gz -- collapsed reads.
• AML025_rep1_vs_ERWOOD_hp.result.count.txt -- shRNA hits.

.
hpreads/
└── AML025
    ├── AML025_rep1.mature.count.txt.gz
    └── AML025_rep2.mature.count.txt.gz
result/
└── batch3
    └── AML025
        ├── AML025_rep1_vs_ERWOOD_hp.result.count.txt
        ├── AML025_rep1_vs_ERWOOD_hp.result.count.txt.all
        ├── AML025_rep1_vs_ERWOOD_hp.result.count.txt.allmature
        ├── AML025_rep2_vs_ERWOOD_hp.result.count.txt
        ├── AML025_rep2_vs_ERWOOD_hp.result.count.txt.all
        └── AML025_rep2_vs_ERWOOD_hp.result.count.txt.allmature

Versioning

We use SemVer for versioning. For the versions available, see the tags on this repository.

Authors

• Yang Shen - Develope the pipeline

Contact

Please contact us if you find bugs, have suggestions, need help etc. You can either use our mailing list or send us an email:

• Yang Shen

shRNAscreen is developed in the Genome Institute of Singapore

License

This project is licensed under the MIT License - see the LICENSE.md file for details