BAM file header after splitting paired end reads

Question

[ilyco]

Hi,

I am processing BAM files using pysam as part of the HiFive set of tools for analysing Hi-C data. The input required are two BAM files with each one representing the forward and the reverse reads respectively from an alignment of paired-end reads.

In order to split the alignment BAM file into forward and reverse reads, I used:

samtools view -f 0x40 file_chr1.bam > file_chr1_1.bam
samtools view -f 0x80 file_chr1.bam > file_chr1_2.bam

Then discovered that the new BAM files do not have headers so I added the -h tag to the commands:

samtools view -h -f 0x40 file_chr1.bam > file_chr1_1.bam
samtools view -h -f 0x80 file_chr1.bam > file_chr1_2.bam

Now I am getting the following error:

for I in range(len(input.header['SQ'])):
  File "pysam/calignmentfile.pyx", line 1486, in pysam.calignmentfile.AlignmentFile.header.__get__ (pysam/calignmentfile.c:16469)
ValueError: malformatted header: no ':' in field

Any ideas on how to fix this?

Thank you very much

[Benn]

Can you show me the header of the bam file?

samtools view -H bamFile.bam

Answer 1

[Pierre Lindenbaum]

You have created SAM files, not BAM files. Use -b to force binary.

[ilyco]

Thank you very much. I used -b and made sure it is a BAM file. Even so, it still would not let me read the file and output "malformatted header".

I then used samtools reheader and removed chromosome fragments e.g. chr_Un from the header.

HiFive still does not recognise the pair of reads and outputs "data is not valid" but at least pysam is reading the BAM files.