Bowtie alignment with two reads per sample

Question

[pel]

The Linux code for running bowtie on our cluster is e.g.:

bowtie -p 20 -S /archive/bowtie_index/cvbnddf sample.fastq sample.sam

However, I have two read files per sample like sample_553_R1.fastq and sample_553_R2.fastq, so what's the bowtie switch setting to use both reads when creating the alignment? I believe I can create a string variable and set equal it to the two variables with a separator(?) and then supply that string in the bowtie option(?)

Answer 1

[Devon Ryan]

bowtie2 -p 20 -x /archive/bowtie_index/cvbnddf -1 sample_R1.fastq -2 sample_R2.fastq -S sample.sam

I assume you really meant bowtie2 rather than the older bowtie1, though the parameters are largely the same. I also hope you're aligning to the transcriptome.

[pel]

Thanks for the syntax for coupling the read files. We have to align to the rat, and I am not sure if the UCSC genome references in bowtie are for the transcriptome(?) I believe you are correct, since the lab director said the data were from mRNAs, and I am not sure how bowtie2(bowtie) delineates RNA-seq data from DNA-seq data (alignments). Is the human reference refseq 19 for DNA and not RNA? You could send a paper link distinguishing DNA from RNA alignments and bowtie2(?)

[Devon Ryan]

If you downloaded the index from iGenomes then this is not a transcriptomic reference (assume this is always the case, regardless of where you download an index). You should use tophat2 or hisat instead. Bowtie2 is not appropriate for alignment of RNAseq data when used alone, that's the point of tophat2.

[pel]

ok, thx for the info!