I have run the MACS,is the outcome right?

[hanifk]

I love this forum very much,everyone here is very kind.

I have run the MACS on my bacterial whose genome is about 26mb
and there are about 10000000 reads mapped to the genome.

However, the out file of MACS contains only 53 peak regions,much less than the 3000 predicted gene.

Maybe there are some problem about my outcome.
my command is
macs14 -t testsorted.bam -f BAM -g 3.0e6 -n test -w
Could you give me some advice?
(my data is rna-seq, I don't have the control file, so I did't use the -c option,is it ok?)

ps.
I also countered this odd problem.
when I run
macs14 -t testsorted.bam -f BAM -g 3.0e6 -n test -w
in the bash,it is ok.

but,when I write the same command in the .sh file,and qsub it.
I will get this error.
/usr/local/bin/python: bad interpreter: No such file or directory

I am not familiar with python,could anyone tell me how to deal with this problem? I have used qsub successfully for many times before today.

[dawe]

I have run the MACS on my bacterial whose genome is about 26mb
and there are about 10000000 reads mapped to the genome.

However, the out file of MACS contains only 53 peak regions,much less than the 3000 predicted gene.

Maybe there are some problem about my outcome.
my command is
macs14 -t testsorted.bam -f BAM -g 3.0e6 -n test -w

if your genome is ~26 Mbp, why do you specify 3.0e6 (and not 2.6e7)?

(my data is rna-seq, I don't have the control file, so I did't use the -c option,is it ok?)

Mmm... What are you trying to do? MACS looks for peaks, if you want to fool MACS treating your RNA as TF enrichments, I believe you have to play with the lambda values, this because the size of your "peak" is probably comparable to the size of the space between peaks (if not the genome itself). I mean, the statistical model could not be the best.
Why don't you try cufflinks?

ps.
I also countered this odd problem.
when I run
macs14 -t testsorted.bam -f BAM -g 3.0e6 -n test -w
in the bash,it is ok.

but,when I write the same command in the .sh file,and qsub it.
I will get this error.
/usr/local/bin/python: bad interpreter: No such file or directory

Check that python is installed in /usr/local/bin on each node of your cluster (or at least the exec node, targeted by qsub).

d