samtools merge

[frymor]

Hallo everybody,

I am running a bowtie assembly for drosophila.
After the bowtie is finished I'm piping the unmapped reads to tophat to see if I can map some more reads onto the same reference genome.

The sam out from bowtie I than convert into bam and tophat make automatically a bam file
After finishing both runs, I would like to combine both bam files with the samtools merge command:

samtools merge -h dilptotal.sam dilptotal_2.bam dilptotal_bowtie.bam dilp_tophat.bam

but I'm keep getting this error message:

[bam_merge_core] different target sequence name: 'YHet' != '2L' in file 'dilp_tophat.bam'

I don't exactly understand what this error means.

I used for both runs the same reference genome. in both there are the chromosomes "2L' and 'YHet'.
YHet is the heterochromatin part of the Y chromosome. It comes 4 times in the sorted bowtie bam file but over 4500 times in the sorted tophat bam file.
'2L' reads I have many millions in both files.

why does it has this problem? Is it because I don't have a header in my tophat output file with the chromosomes (@SQ)?

can I set tophat to have an header in the sam or bam files?

Thanks for ant advice,

Assa

[nilshomer]

Make sure that the SQ lines in the header are the same (use samtools view -H). You will also need to sort them before merging (samtools sort).

[frymor]

Make sure that the SQ lines in the header are the same (use samtools view -H). You will also need to sort them before merging (samtools sort).

This is exactly my problem. tophat produces no header in the bam file.

Can I change the setting so that tophat will create a header?
Is there a header in the (temporary) sam files from tophat?

Is it enough just to copy paste the header from the bam file from bowtie into the one from tophat?

[hbt]

This is exactly my problem. tophat produces no header in the bam file.

Can I change the setting so that tophat will create a header?
Is there a header in the (temporary) sam files from tophat?

Is it enough just to copy paste the header from the bam file from bowtie into the one from tophat?

I'd love to know the answers to this too!

cheers

[ozs2006]

I got similar problem:

I downloaded sam files from recent published study.
Each sam file contains alignments of the reads to a single chromosome (hg19).
I want to merge alignments into one file.
Every sam file have only @SQ as header of its chromosome.

For example:
in chrY.sam
@SQ SN:chrY LN:59373566
in chrM.sam
@SQ SN:chrM LN:16571

I used:
samtools view -T /data/pipeline_in/Genomes/Human_GRCh37/all.fa -Sb chrY.sam | samtools sort - chrY.sam.sorted
samtools view -T /data/pipeline_in/Genomes/Human_GRCh37/all.fa -Sb chrM.sam | samtools sort - chrM.sam.sorted

Then in order to merge them:
samtools merge out chrM.sam.sorted.bam chrY.sam.sorted.bam

I got this error:
[bam_merge_core] different target sequence name: 'chrM' != 'chrY' in file 'chrY.sam.sorted.bam'

What I need to do?
from searching the net I got some clues this error is connected to the header?
Do I need to replace the headers of the primary sam files?
Where I find proper example for header?

Thanks in advance,
Oz Solomon