paired-end, bam-file-format

[Azazel]

1.) Having a bam-file of paired-end data, how can I find out what the pairs are? I.e. which sequence is the start and which one the end of which pair?

2.) If I only want the 5-prime end of each pair: how can I remove all the downstream pairs?

Thanks!

[maubp]

Look at the FLAG bit field.

[Azazel]

Look at the FLAG bit field.

0x1 template having multiple fragments in sequencing
0x2 each fragment properly aligned according to the aligner
0x4 fragment unmapped
0x8 next fragment in the template unmapped
0x10 SEQ being reverse complemented
0x20 SEQ of the next fragment in the template being reversed
0x40 the rst fragment in the template
0x80 the last fragment in the template
0x100 secondary alignment
0x200 not passing quality controls
0x400 PCR or optical duplicate

I do not understand the semantics of this and anyway I guess it's a little bit hard working with the bam-files directly anyway. I mean how are you supposed to look at the bit-field, with a hex-editor?!

Is there an easier way? For example, can't I convert to bed-format, so that the paired-end information is retained? I tried bamToBed, but there as well I don't see how the pairs are matched. I also tried "view" from samtools.

[maubp]

Try samtools view with -X to explain the flags.

[Martino]

Use Pysam in Python!

You can use Pysam in Python.

First, you have to install this library.

After that, you have to type the following code:
====================================

import pysam
samfile = pysam.Samfile("ex1.bam", "rb")
pairedreads = pysam.Samfile("allpaired.bam", "wb", template=samfile)
for read in samfile.fetch():
if read.is_paired:
pairedreads.write(read)

pairedreads.close()
samfile.close()
=====================================

This code is from: http://wwwfgu.anat.ox.ac.uk/~andreas...tools/api.html

There, you can read more about it.

The function read.is_paired looks, if a read is a paired-end read or a single read.

I hope this helps you...