NGSclean

A pipeline to trim the reads and remove ribosomal RNA from RNA-Seq data

This pipeline trims the adaptor from the ends of reads, and move plant Ribosomal reads from RNAseq reads. (This tool also works for data from animal species, when corresponding rRNA sequences is provide.) The process is based on two tools, Trimmomatic and STAR for trimming and mapping respectively.

First of all please make sure how to run these two tools on your system.

On Sapelo/UGA, it is like this:
Trimmomatic

module load ava/jdk1.8.0_20 
/usr/local/apps/trimmomatic/0.33/trimmomatic-0.33.jar 

STAR

modle load java/jdk1.7.0_67
/usr/local/apps/star/latest/bin/STAR

Dependency:
Trimmomatic http://www.usadellab.org/cms/?page=trimmomatic
STAR https://github.com/alexdobin/STAR

0. Copy the scripts and generate index file for rRNA sequences

copy NGSclean directory to your system, for example
/lustre1/lxue/NGSclean

prepare rRNA reference for STAR

NGSclean=/lustre1/lxue/NGSclean 

cd $NGSclean
mkdir rRNA_ref
module load java/jdk1.7.0_67
/usr/local/apps/star/2.4.2a/bin/STAR \
  --runThreadN 2  \
  --runMode genomeGenerate  \
  --genomeDir rRNA_ref  \
  --genomeChrBinNbits  5 \
  --genomeFastaFiles rRNA_only_NR.fas  

Keep '--genomeDir rRNA_ref' as it is change 'rRNA_only_NR.fas' if a new reference file is necessary.

1. Prepare design file

cd /lustre1/lxue/RNAseq/01clean  
NGSclean=/lustre1/lxue/NGSclean  
python $NGSclean/generate_design_file.py -f /lustre1/lxue/RNAseq/00reads -d RNAseq_design.txt -p 

Check the design file. Revise the sample names if necessary, it will be used as a prefix for fastq files
more RNAseq_design.txt 

2. Trim and Clean

cd /lustre1/lxue/RNAseq/01clean
settings
NGSclean=/lustre1/lxue/NGSclean
trimmoFull=/usr/local/apps/trimmomatic/0.33/trimmomatic-0.33.jar 
starFull=/usr/local/apps/star/latest/bin/STAR
adaptor=/usr/local/apps/trimmomatic/latest/adapters/TruSeq3-PE.fa
trimmo_module=java/jdk1.8.0_20 
star_module=java/jdk1.7.0_67

generate shell files

-s how to treat the singleton reads:merge, keep , discard
-q queue to run the jobs: queue(defualt), inter(run interactive jobs)

python $NGSclean/trim_and_clean.py -d RNAseq_design.txt -t 8  -s merge \
  --run_trimmomatic $trimmoFull --load_trimmo_module $trimmo_module  --adaptor $adaptor  \
  --run_star $starFull --load_star_module $star_module 

python $NGSclean/trim_and_clean.py -d RNAseq_design.txt -t 8  -s keep \
  --run_trimmomatic $trimmoFull --load_trimmo_module $trimmo_module  --adaptor $adaptor  \
  --run_star $starFull --load_star_module $star_module 

python $NGSclean/trim_and_clean.py -d RNAseq_design.txt -t 8  -s discard -q inter \
  --run_trimmomatic $trimmoFull --load_trimmo_module $trimmo_module  --adaptor $adaptor  \
  --run_star $starFull --load_star_module $star_module 

And then run the master shell script to submit the jobs or run them interactively.
For example:

chmod 750 RNAseq_design.sh
./RNAseq_design.sh