Samtools bam sort and index errors

[ddmoore1028]

I'm having issues with the sorting/indexing of a bam file. I just started working with samtools (and all bioinformatics) last week so I'm pretty new to this.

First I generated a SAM alignment file using STAR. Then I converted that to a BAM file and it seemed to run correctly. But the I try to sort the bam file, and here are my inputs and the resulting output:

samtools sort alignment1.bam alignment1.sorted
[bam_header_read] EOF marker is absent. The input is probably truncated.
[bam_header_read] invalid BAM binary header (this is not a BAM file).
Segmentation fault

If I try to go straight to indexing, I get a similar error:

samtools index alignment1.bam alignment1.bai
[bam_header_read] EOF marker is absent. The input is probably truncated.
[bam_header_read] invalid BAM binary header (this is not a BAM file).
[bam_index_core] Invalid BAM header.[bam_index_build2] fail to index the BAM file.

Can someone help me fix these issues?

Thanks

[GenoMax]

For the first part see this thread (post #2): http://seqanswers.com/forums/showthread.php?t=21908

Are you doing this using a virtual (or real) server?

[Richard Finney]

Show your command for creating the BAM from the SAM.
What version of samtools are you using?

If you know the "od" command, try this ...
od -x YOURFILENAME.bam | head -1

output should be

0000000 8b1f 0408 0000 0000 ff00 0006 4342 0002

The message you are getting is from this code which is checking for the "BAM\001" characters ...

magic_len = bam_read(fp, buf, 4);
if (magic_len != 4 || strncmp(buf, "BAM\001", 4) != 0) {
fprintf(stderr, "[bam_header_read] invalid BAM binary header (this is not a BAM file).\n");
return 0;
}

Possible reasons are : the file zero length, the file is a sam file, it is corrupted.

Can you view the first few bytes of the "bam" file in question ?

[ddmoore1028]

Are you doing this using a virtual (or real) server?

I'm not entirely sure. I'm accessing a computer that I think has 24 cores using the ssh command in Unix, and using that to run my programs. My program files and data are stored in what I think is probably a virtual server (to access it directly from my macbook, not using Unix, I use the Go > Connect to server function in the Finder).

Show your command for creating the BAM from the SAM.

nohup samtools view -bS Aligned.out.sam > alignment1.bam &>secondtrybam.out&

I tried the od command and got this:

od -x alignment1.bam | head -1
0000000

My program file is : samtools-0.1.19

Can you view the first few bytes of the "bam" file in question ?

I just tried to use vi, more, and less to view and there doesn't seem to be anything in the file. Is there something wrong with my SAM>BAM conversion?

Thank you

[ddmoore1028]

When I checked the SAM file, it definitely seems correct. The first number of lines begin with @ and seem to refer to the chromosomes mapped, and after out I see line after line of reads

[Richard Finney]

What does "&>" do?

Try running just this ...
samtools view -bS Aligned.out.sam > alignment1.bam

What happens ?
What is the size of alignment1.bam ?

[ddmoore1028]

Since nohup makes the program run in the background and not quit if I have to log out of my laptop, I thought "&>" changed the nohup.out file to the name I assigned it, but I think that may have changed the command somehow.

The original bam file that was giving me trouble is 2 kb.

The new bam file is 622 mb and counting. When I entered the simpler command I got:

samtools view -bS Aligned.out.sam > alignment1again.bam
[samopen] SAM header is present: 25 sequences.
[]

and its been running since. I think this is probably going to fix it! Thank you!

The SAM file is about 50 gb, what can I expect the bam file size to be at the end?

[ddmoore1028]

I checked the size of the secondtrybam.out file that I used &> on, and it's 13.04 gb. I think that the &> directed the binary to write to the .out file rather than the .bam file. I think that's the problem I had.