You can just replace "> ${filename}_chr.bam" with "> /new/path/${filename}_ch.bam".

[a_l_25]

stack in the output part

how to put the filename variable in different path than the original path where bam files located. Hence , keeping the variable $filename as part of the every new bam file's name.

The following line works well for me. It will replace the names of chromosome 1 to 22, X, Y and MT (mitochondrial DNA). It will create a new file with the name test_chr.bam instead of test.bam

use Samtools
samtools view -H test.bam | sed -e 's/SN:\([0-9XY]\)/SN:chr\1/' -e 's/SN:MT/SN:chrM/' | samtools reheader - test.bam > test_chr.bam

If you want to do this for a bunch of bam files at once, use a loop. The variable $filename is your file without the extension (assuming there is no period except the one in ".bam").

for file in *.bam; do filename=`echo $file | cut -d "." -f 1`; samtools view -H $file | sed -e 's/SN:\([0-9XY]\)/SN:chr\1/' -e 's/SN:MT/SN:chrM/' | samtools reheader - $file > ${filename}_chr.bam; done

[neurongs]

thank you mra for this useful tool!

[mra]

BAM to BAM plus "chr"

The chromosome names can be changed on the fly. This command takes an input BAM file with chr names '1, 2, 3,..,X, Y, MT' and output a BAM file with names 'chr1, chr2, chr3,..., chrX, chrY, chrM' on both read and header lines. The output BAM doesn't include any non-canonical chromosome that could be present in the input BAM (such as JG*, Un*, etc).

samtools view -h INPUT.bam | awk 'BEGIN{FS=OFS="\t"} (/^@/ && !/@SQ/){print $0} $2~/^SN:[1-9]|^SN:X|^SN:Y|^SN:MT/{print $0}  $3~/^[1-9]|X|Y|MT/{$3="chr"$3; print $0} ' | sed 's/SN:/SN:chr/g' | sed 's/chrMT/chrM/g' | samtools view -bS - > OUTPUT.bam

[dpryan]

my header.sam file looks OK. "chr" has been added.

But when I run the samtools reheader command, nothing changes in the original .bam file...

The reheader command doesn't change the original file (I'm not surprised you thought it would, the name lends itself to that belief and I don't think the documentation mentions this). You'll need to output things to a file rather than to samtools view. Try something like:

samtools reheader header.sam sample1.bam > sample1.fixed.bam

[ddaneels]

my header.sam file looks OK. "chr" has been added.

But when I run the samtools reheader command, nothing changes in the original .bam file...

Code:
samtools reheader header.sam sample1.bam | samtools view -H sample1.bam

Sample1.bam is my original file, so I was hoping that the header in sample1.bam would have changed, but it didn't.

I'm a programming newbie ... so maybe there's a mistake in my code?

[maubp]

OK - I may have been worrying over nothing then.

[dpryan]

I don't think that would work. Using 'samtools reheader' would only edit the embedded SAM header embedded in a BAM file, it would not IIRC update the separate BAM specific header table containing the list of references (their names and references).

You could turn the BAM file into SAM (e.g. with samtools view -h), do the replacement (e.g. with sed), and then optionally convert back to BAM (again with samtools view). That can be done as one line by piping the output from one tool to the next.

Actually, bam_reheader runs the full bam_header_write using only the new header, so it seems it does both (I haven't bothered looking into the source of bam_header_write, I should note). I decided to run a quick test, since I can't say I've ever actually run the reheader command. For that, I took the header of a sorted alignment (written to a file called header.sam), and changed "chr1" to "chr100".

samtools view accepted_hits.bam | head -n 2
HWI-ST143:530:C102UACXX:5:1101:3568:162900	272	chr1	3005607	0	51M	*	0	0	CATAAATTCATTTTTTAATAGCTGAGTAGTATTCCATTGTGTAAATGTACC	*	AS:i:0	XN:i:0	XM:i:0	XO:i:0	XG:i:0	NM:i:0	MD:Z:51	YT:Z:UU	NH:i:20	CC:Z:=	CP:i:105668244	HI:i:0
HWI-ST143:530:C102UACXX:5:1308:5464:137163	272	chr1	3006556	0	51M	*	0	0	TTAGCTCCCTTGTCAAAGATCAGGTGACCATAGGTGTGTGGATTCATCTCT	*	AS:i:0	XN:i:0	XM:i:0	XO:i:0	XG:i:0	NM:i:0	MD:Z:51	YT:Z:UU	NH:i:20	CC:Z:chr15	CP:i:16439024	HI:i:0

samtools reheader header.sam accepted_hits.bam | samtools view - | head -n 2
HWI-ST143:530:C102UACXX:5:1101:3568:162900	272	chr100	3005607	0	51M	*	0	0	CATAAATTCATTTTTTAATAGCTGAGTAGTATTCCATTGTGTAAATGTACC	*	AS:i:0	XN:i:0	XM:i:0	XO:i:0	XG:i:0	NM:i:0	MD:Z:51	YT:Z:UU	NH:i:20	CC:Z:=	CP:i:105668244	HI:i:0
HWI-ST143:530:C102UACXX:5:1308:5464:137163	272	chr100	3006556	0	51M	*	0	0	TTAGCTCCCTTGTCAAAGATCAGGTGACCATAGGTGTGTGGATTCATCTCT	*	AS:i:0	XN:i:0	XM:i:0	XO:i:0	XG:i:0	NM:i:0	MD:Z:51	YT:Z:UU	NH:i:20	CC:Z:chr15	CP:i:16439024	HI:i:0

So, it seems to work.

[maubp]

Look at the samtools reheader command. Presumably you would just "samtools view -H file.bam > header.sam", edit the header, and then use that with reheader.

I don't think that would work. Using 'samtools reheader' would only edit the embedded SAM header embedded in a BAM file, it would not IIRC update the separate BAM specific header table containing the list of references (their names and references).

You could turn the BAM file into SAM (e.g. with samtools view -h), do the replacement (e.g. with sed), and then optionally convert back to BAM (again with samtools view). That can be done as one line by piping the output from one tool to the next.

[maubp]

You are mistaken. The reads only contain a number that tells which entry from the header to pick.

Or to be more precise, in BAM the reads just store an integer to say which reference it mapped to (referencing a table at the start of the BAM file, which is separate to any embedded SAM header), but in SAM the reads store the reference sequence's name.

[xied75]

1, This Big Huge Black 1, is column 3, not 7. It is not a number, but a string.
2, You don't need to change this 1 to chr1, is because programs like BWA and Samtools will read both format; but it doesn't mean this is a number ref to the header lines. Your understanding is more correct.

[ddaneels]

Thanks for the info.

Everything will work fine then. I was confused with the "1" in the 7th column of the read-part in the .bam file.

Example:

HWI-ST571_103:4:1302:9610:62449 99 1 604269 254 100M = 60324 152 GGAA...

I thought that the highlighted 1 also had to be changed to chr1.

[xied75]

You are mistaken. The reads only contain a number that tells which entry from the header to pick.

In SAM spec v 1.4 document, column 3 RNAME is of String type, to hold Reference sequence NAME of the alignment.

[dpryan]

That would be a way to change the header, but if I'm not mistaking the chromosome numbers are also in the actual read-part of the file. So how to change them there?

As was mentioned by ffinkernagel, this is incorrect. I should note that you should avoid swapping the order of chromosomes or any other major edits. Just adding or removing "chr" won't break anything, but changing the order of things in the header or removing chromosomes could cause issues.