Is there a difference between a .sam produced by tophat2 and bowtie2?

[meabh]

I hope this makes sense,

So I have a perl script that is written to take the accepted_hits.bam file from tophat2, in sam format as an input. Unlike the original paper, I need the unmapped reads as a fastq file so I'm using bowtie2 to map, as it has this option also it outputs mapped results as a sam file so I don't need to convert it. My question is do these files have the same columns or not? It would seem fruitless to try this if tophat2 and bowtie2 output different columns etc.



Thanks

Edit. I don't know how I missed in the tophat manual that you can use the --no-convert-bam to output the result as sam, which means I don't need to use bowtie2 instead

[Michael.Ante]

Hi meabh,

First, the main difference is that TopHat2 is designed to detect and report spliced reads. Whenever there is s a read spanning two exons, you'll see that with TopHat2 alignment. The aligner flags that read in a certain way (e.g. adding the XS attribute). Also the NH attribute (number if hits per read) is set in TopHat2 but not in Bowtie2. Nevertheless, if a read origins from within an exon the two programs will align it to the same location.

In order to convert a bam file into a sam file you can just use samtools view. And to convert a bam file into a fastq file, you can use the Tophat2 script bam2fastx.

Cheers,

Michael