Bowtie output

[statsteam]

Hi all,

I got an RNA seq file that contains about 15million reads.
As it has more than 32bp reads, I decided to use Bowtie instead of my usual pick which is Eland.

So I mapped the sequence on the hg18. For the comparison purpose, I also ran Eland using first 32bp of the original read.

As usual, Eland output showed about 92% mapping rate but the Bowtie output contains only 5.1 million rows which is only about 34% mapping rate.

The options I used when running Bowtie was -f and -t.
The exact command line was:

bowtie -f -t bowtieIndexPath inputRNASequence outputFileName

The Bowtie output contiains all the sequences (5.1 million sequences) mapped with at most two mismatches.
Meanwhile, the Eland mapped 11.2 million sequences with at most two mismatches using first 32bp reads.
Since there were 15 million reads initially,
the mapping rate of Eland with at most two mismatches is 11.2/15 = 0.747
the mapping rate of Bowtie with at most two mismatches is 5.1/15 = 0.340

Is there any way to make Bowtie display the result of each read even though it does not have match?

Thank you,
Statsteam

[krobison]

There is an option in bowtie to trim the sequences - -to get a really fair comparison you could try bowtie with that option set so it is looking only at the first 32 nt.

It would seem unlikely that splice junctions explain things, but you could use TopHat which is the bowtie-based spliced aligner