sam to bam conversion error, no @SQ lines in the header, missing header?

[efoss]

I have a sam file that I want to convert to a bam file:

samtools view -b -S in.sam > out.bam

I get this error:

[samopen] no @SQ lines in the header.
[sam_read1] missing header? Abort!

My sam file looks perfectly normal to me. Does anyone see what I'm doing wrong?

Thanks.

Eric

[efoss]

To add to my previous post: I also tried the following command but I got the same error message:

samtools view -h -b -S in.sam > out.bam

[BAMseek]

I think you might need to specify a reference file containing the names of the sequences and the total sequence lengths. This is done with the "-t" flag. Here is a description from samtools:

-t FILE
This file is TAB-delimited. Each line must contain the reference name and the length of the reference, one line for each distinct reference;

[maubp]

Or from memory -T can specify a FASTA file of the references.

[swbarnes2]

Originally Posted by maubp

Or from memory -T can specify a FASTA file of the references.

samtools view doesn't require -T. It works fine without a reference fasta.

Maybe your headers aren't right?

Here's what mine look like:

@SQ SN:SNP_8787 LN:161
@SQ SN:vector LN:13078

[efoss]

Originally Posted by swbarnes2

samtools view doesn't require -T. It works fine without a reference fasta.

Maybe your headers aren't right?

Here's what mine look like:

@SQ SN:SNP_8787 LN:161
@SQ SN:vector LN:13078

Yes - it was the headers. I see that when I made my sam file, I didn't use the -h option, so then when I went to convert it back to a bam file, I couldn't fix things simply by adding the -h option. I had to go back and remake the sam file with -h and then convert it back to a bam file with -h, and now everything is working fine. Thanks very much, everyone, for the help.

Eric

[prasadg]

Originally Posted by efoss

Yes - it was the headers. I see that when I made my sam file, I didn't use the -h option, so then when I went to convert it back to a bam file, I couldn't fix things simply by adding the -h option. I had to go back and remake the sam file with -h and then convert it back to a bam file with -h, and now everything is working fine. Thanks very much, everyone, for the help.

Eric

Hey efoss,

I am also facing the same problem. I couldn't undertsood this line of yours "I didn't use the -h option, so then when I went to convert it back to a bam file, I couldn't fix things simply by adding the -h option. I had to go back and remake the sam file with -h" how to did you make same file with -h? Did you use Bwa to create sam file.?

[efoss]

Originally Posted by prasadg

Hey efoss,

I am also facing the same problem. I couldn't undertsood this line of yours "I didn't use the -h option, so then when I went to convert it back to a bam file, I couldn't fix things simply by adding the -h option. I had to go back and remake the sam file with -h" how to did you make same file with -h? Did you use Bwa to create sam file.?

Hi prasadg,

It's been a while, but I think that I created a sam file with samtools' view command and that that command accepts a -h flag, which will create a sam file with appropriate headers.

Eric

[prasadg]

Originally Posted by efoss

Hi prasadg,

It's been a while, but I think that I created a sam file with samtools' view command and that that command accepts a -h flag, which will create a sam file with appropriate headers.

Eric

Okay . I will look into it . I have the sam file created by bwa. So I was using samtools to convert sam to bam but I am getting same answer as you have got

[samopen] no @SQ lines in the header.
[sam_read1] missing header? Abort!

So I wanted to know how did you do it .

Thanks

[efoss]

Originally Posted by prasadg

Okay . I will look into it . I have the sam file created by bwa. So I was using samtools to convert sam to bam but I am getting same answer as you have got

[samopen] no @SQ lines in the header.
[sam_read1] missing header? Abort!

So I wanted to know how did you do it .

Thanks

Hi prasadg,

Another thing to try if that doesn't work is using the -r option in bwa to add read group information when you create your original alignment file.

Eric

[prasadg]

Originally Posted by efoss

Hi prasadg,

Another thing to try if that doesn't work is using the -r option in bwa to add read group information when you create your original alignment file.

Eric

I will give it a shot. by -r you mean -R ? cause as I see the alignment options -r is with uppercase. There is no lower case 'r' option.

[efoss]

Originally Posted by prasadg

I will give it a shot. by -r you mean -R ? cause as I see the alignment options -r is with uppercase. There is no lower case 'r' option.

No - I mean -r. Look under sampe and you'll see the option.

Eric

[prasadg]

Originally Posted by efoss

No - I mean -r. Look under sampe and you'll see the option.

Eric

Thanks alot . It certainly made my day. I certainly was on wrong track .

Really appreciate you replys.

[bbm]

I converted sam file to bam, then removes PCR duplicates, then convert bam to sam. It showed the same error msg:

error msg: [samopen] no @SQ lines in the header.
[sam_read1] missing header? Abort!

I tried -h option, still didn't work...

[efoss]

Hi bbm,

You might want to try Picard's AddOrReplaceReadGroups:

http://picard.sourceforge.net/comman...laceReadGroups

[Risha]

I also came across the 'missing header' error. I found that that when converting BAM-> SAM, I had forgotten about printing the headers. This fixed it for me:


BAM -> SAM:

Note : -h print header for the SAM output

samtools view -h in.bam > out.sam

SAM -> BAM (even SAM produced from BWA):

samtools view -bS in.sam > out.bam

[Lovro]

i have the same problem...

this is from the man samtools:
o Import SAM to BAM when @SQ lines are present in the header:

samtools view -bS aln.sam > aln.bam

If @SQ lines are absent:

samtools faidx ref.fa
samtools view -bt ref.fa.fai aln.sam > aln.bam

where ref.fa.fai is generated automatically by the faidx command.

However, i don't know which ref.fa to use.
The sam i'm trying to convert to bam, sort and index was generated by a foreign seq searching program, readscan, which uses host and pathogen ref seqs and try to map to both of them.

the actual sam file i'm trying to convert is "ambiguous.sam" which contains reads mapped to both ref seqs.

can somebody please explain what is going on... what are @SQ lines, why does samtools need them to convert to bam and index the file, why am i expected to provide a ref.fa ....

its really confusing

tnx !

[mastal]

If you are aligning to something like the human genome, then you would usually have one @SQ line for each chromosome in the sam file header.

If you have the sequences of the host and pathogen the reads were mapped to, you could combine them into a ref.fa file, and run samtools faidx.