RNA-seq Pipeline for Known Transcripts

Sequence Quality and Trimming

1. Run FASTQC to assess quality of reads from sequencer and:
2. FASTQC available at http://www.bioinformatics.bbsrc.ac.uk/projects/fastqc/
   1. Open run_fastqc on a windows machine. Individually open each sequence file and allow it to analyse. Save this report.
   2. Check this report to decide if sequences need to be trimmed or discarded.

Filter Sequences Using FastX-Toolkit

1. Filter for quality, if applicable
2. Trim, if applicable

Generate a Reference Genome

1. Run bowtie-build to generate Burroughs Wheeler transformed reference genome (.ebwt format).
2. http://bowtie-bio.sourceforge.net/index.shtml (bowtie, tophat, and cufflinks are here).
3. [Optional input and parameter settings are in square brackets.]
4. <Required parameters are in greater than/less than brackets.>
5. This BW transformed reference genome can be created once then used repeatedly in the future.
6. $ is the command prompt.

$ bowtie-build [-f specifies reference genome is in fasta format] <path to input reference genome (e.g. /ccmb/CoreBA/BioinfCore/Common/DATA/BowtieData/H_Sapiens/hg19.fa)> <base name for reference genome output .ebwt files (e.g hg19)>

Align Reads to Reference Genome with Tophat

Run tophat to align reads to the reference genome. I’ve included a pseudo command line as well as a “real” command line.

$ tophat [-p #processors -o ./output_directory] <./reference genome in both .ebwt and fasta formats (e.g. /ccmb/CoreBA/BioinfCore/Common/DATA/BowtieData/H_Sapiens/hg19)> <reads file to be aligned (e.g. s_1_1_sequence.fastq)>
$ tophat -p 5 -o ./HG19/tophat_out_hg19_001_trimmed /ccmb/CoreBA/BioinfCore/Common/DATA/BowtieData/H_Sapiens/hg19 ./HG19/Rich_trim/A_1_16_85.fastq

Use Cuffcompare to Generate .gtf Reference

Run cuffcompare to create .gtf format reference genome from a generic reference genome. Note that cuffcompare adds the tss_id and p_id columns that you will need in cuffdiff. This .gtf reference can be created once then used repeatedly in the future.

$ cuffcompare [-o ./output_directory] < input file twice (e.g. /ccmb/CoreBA/BioinfCore/Common/DATA/CufflinksData_hg19/hg19.gtf /ccmb/CoreBA/BioinfCore/Common/DATA/CufflinksData_hg19/hg19.gtf )>

$ cuffcompare -o ./cuffcompare_out /ccmb/CoreBA/BioinfCore/Common/DATA/CufflinksData_hg19/hg19_genes.gtf /ccmb/CoreBA/BioinfCore/Common/DATA/CufflinksData_hg19/hg19_genes.gtf

Use Cuffdiff to Identify Differentially Expressed Transcripts

Run cuffdiff to identify differentially abundant transcripts.

$ cuffdiff  [-p #processors -o ./output_directory –L label1,label2,etc. –T (for time series data) –N (use upper quantile normalization –compatible_hits_norm (use reference hits in normalization) –b (use reference transcripts to reduce bias, include path to file e.g. /ccmb/CoreBA/BioinfCore/Common/DATA/BowtieData/H_Sapiens/hg19.fa) –u (improve multi-read weighting) ] <transcripts.gtf (produced by cuffcompare) sample_A_accepted_hits1.bam, sample_A_accepted_hits2.bam,etc (all produced by tophat) sample_B_accepted_hits1.bam,sample_B_accepted_hits2.bam, etc> 

$ cuffdiff -o ./HG19/Cuffdiff_out_options_b_u_N_compatible/ -p 14 -L Control,PUF_kd --no-update-check -b /ccmb/CoreBA/BioinfCore/Common/DATA/BowtieData/H_Sapiens/hg19.fa -u -N --compatible-hits-norm /ccmb/CoreBA/BioinfCore/Projects/Goldstrohm_McEachin/HG19/cuffcompare_out.combined.gtf /ccmb/CoreBA/BioinfCore/Projects/Goldstrohm_McEachin/HG19/tophat_out_hg19_001_trimmed/accepted_hits.bam /ccmb/CoreBA/BioinfCore/Projects/Goldstrohm_McEachin/HG19/tophat_out_hg19_002_trimmed/accepted_hits.bam