Pipeline for RNA-Seq data analysis

The pipeline is for RNA-Seq data analysis on condo, located in:

/work/LAS4/xfzhao/tutorial/RNA-Seq/ensembl_gm01

for gsnap and tophatp.

The configuration file for the pipeline is proj.cfg in:

ensembl_gm01/tophat(gsnap)-cufflinks-htseq

Steps to run the pipeline

Transfer your data your home on /work/LAS/your-lab

Do a:

  • md5sum check if needed
  • fastqc if needed.
  • clean the reads if needed.

Select the pre-built indexed database for your species in:

/ptmp/LAS/BioDatabase/iGenomes

or use your own indexed database.

Pre-build Indexed Databases are:

/ptmp/LAS/BioDatabase/iGenomes/Species/DataSource/BuildVersion/Sequence/Pre-IndexedDatbase

Species: commonly anaylized ones. Add other species upon request.
DataSource: NCBI, Ensembl and UCSD
BuildVersion: last three builds are online
Pre-indexed Database: BWAIndex, Bowtie2Index, BowtieIndex, GmapIndex

The annotation file is in:

/ptmp/LAS/BioDatabase/iGenomes/species/DataSource/build/Annotation

Edit proj.cfg for RNA-Seq pipeline

The sample data is for Glycin Max, sequenced on 3 lanes for each sample, and combined together during alignment using gsnap or tophat.
The annotation is Gm01 from Ensembl. After the alignment, the raw read was counted using HTseq-count, and FPKM was calculated by cufflinks.
The tab-delimited FPKM/raw count summary files are created using the script summary_fpkm_ct.sh and summary_htseq_ct.sh for
the down-stream statistical analysis.

Example : RNA_Seq analysis using GSNAP, HTseq-count, and cufflinks.

Copy the:

/work/LAS4/xfzhao/tutorial/RNA-Seq/ensembl_gm01/gsnap-cufflinks-htseq

to your home. Edit the project configuration file:

gsnap-cufflinks-htseq/proj.cfg

The meaning of the parameters is self explanatory.

run gsnap

cd gsnap-cufflinks-htseq/aln

# check up gsnap\_align\_.sh

qsub parallel-aln.qsub

run HTSeq-count

cd gsnap-cufflinks-htseq/htseq

check gsnap_HTseq_count_.sh

then submit parallel-HTseq_ct.qsub.

The scripts summary_htseq_ct.sh and summary_htseq_ct_run.sh are for making the raw count summary file.

cd gsnap-cufflinks-htseq/clout,

check gsnap_cufflinks_.sh.

then submit parallel-cufflinks.qsub to run the job.
The scripts summary_fpkm_ct.sh and summary_fpkm_ct_run.sh are for making the tab-delimited FPKM summary file.