rnaseq_scripts

Custom scripts and tools for running the guigolab/grape-nf pipeline on Amazon EC2 (Amazon Linux AMI) or other CentOS/RedHat machines

Requirements

• System running CentOS 6.7/7 or Amazon Linux AMI
• at least 8 cores
• at least 32GB RAM
• approximately 50GB of free disk space per raw data file

Installation

Run the following command in a shell on the machine where you want to process your data

curl -fsSL https://github.com/leshaker/rnaseq_scripts/raw/master/install_rnaseq_pipeline.sh | bash

The script will install all dependencies and tools needed for processing files from GEO, CCLE or other sources (.bam, .sra or .fastq files).

For installing only the sudo (packages, docker, tools in /opt/) or the user part (pipeline, reference genome etc.), type

curl -fsSL https://github.com/leshaker/rnaseq_scripts/raw/master/install_rnaseq_pipeline_sudo.sh | bash

or

curl -fsSL https://github.com/leshaker/rnaseq_scripts/raw/master/install_rnaseq_pipeline_user.sh | bash

respectively.

Processing data from GEO

Add data sets to the file GEO_data.txt in the following format

SRR2537160 GSM1898288_polycysticstemcell_expansionmedium_1_17p6

where the first part represenst the SRA run identifier from SRA and the second is the filename (ideally containing the GEO or SRA identifier).

Then run the script run_loop.sh for downloading, converting and processing all files in the GEO_data.txt file.

cd ~/RNAseq_pipeline
./run_loop.sh GEO grape

Consider running the command within a screen as the processing will take about 4h per file (on 36 core, 60GB RAM machine).

Processing data from CCLE

Add data sets to the file CCLE_data.txt in the following format

b39b60cd-ed66-4824-9548-6e1396da753c	G20463.C2BBe1.2.bam
e6b5d8f8-76ac-4598-954a-aadbf4306afa	G27383.CL-40.1.bam

where the first part represenst the Analysis Id and the second is the Filename from the CGHub Browser

Then run the script run_loop.sh for downloading, converting and processing all files in the CCLE_data.txt file.

cd ~/RNAseq_pipeline
./run_loop.sh CCLE grape

Consider running the command within a screen as the processing will take about 4h per file (on 36 core, 60GB RAM machine).

Processing user supplied data

Add data sets to the file USER_data.txt in the following format

NK0_rep1	Sample1_NK_cells_untreated
NK0_rep2	Sample2_NK_cells_untreated
NK0_rep3	Sample3_NK_cells_untreated
NK5_rep1	Sample1_NK_cells_treated_with_5mg
NK5_rep2	Sample2_NK_cells_treated_with_5mg
NK5_rep3	Sample3_NK_cells_treated_with_5mg

where the first part represents the input filename (withouth fastq.gz extension) and the second is the output filename.

Then run the script run_loop.sh for processing all files in the USER_data.txt file.

cd ~/RNAseq_pipeline
./run_loop.sh USER kallisto

Consider running the command within a screen as the processing will take about 2h per file (on 36 core, 60GB RAM machine).