Bowtie maps more ChIP-Seq reads than Tophat

[tulia]

Hi,

I am mapping a several sets of ChIP-Seq data and I am getting more reads mapped with Bowtie than with Tophat. This doesn't make sense to me.

My bowtie command:
./bowtie-0.12.7/bowtie -S --fr -p 6 --un unaligned.fastq \
mm9 \
sample.fastq \
accepted_hits.sam 2>filter.bowtie.err

/share/apps/samtools/samtools view -Sb -F 4 accepted_hits.bam > mapped.bam

Reads mapping to reference eg: 12,151,483


My tophat command:
tophat-1.4.1.Linux_x86_64/tophat \
-p 6 \
--library-type fr-unstranded \
-G genes.gtf \
-o tophat \
--no-coverage-search \
mm9 \
sample.fastq

Reads mapping to reference eg:10,392,434

Any thoughts on why this could be?

[dpryan]

Why do you find that odd? Tophat is meant for RNAseq and runs bowtie with a number of changed options. That it produces different and worse results for ChIP-seq should be expected.

[tulia]

Thanks for responding. I was guess I was expecting it to behave like RNA-Seq data and produce better results with Tophat than Bowtie because of the way the mapping is performed.