samtools mpileup hates Bowtie2???

[peromhc]

Hi All,

I have been struggling with an issue where mpileup refuses to work find variation in a given bamfile. In fact, when using the standard mpileup routine described on the samtools website, the vcf file contains only 11 'SNPs', all of which are filtered out by the default conditions in vcfutils.pl step. This is a pooled sample from multiple unrelated individuals (birds), so I know that there is variation.

Anyway, I had been using Bowtie2, but on a whim, tried Bowtie1, and like magic, now there are 10K SNP's, which is a reasonable number for my study design.

So, I am wondering, what is it about Bowtie2 that mpileup does not like?

Here is a SAM entry for a read in bowtie2

HS2:124:C097KACXX:7:2307:12782:66337	113	4	2	42	98M	396128	3022	0	AGCTCCAGCCATTCCCTGGGGGCTGTCCCTGATCATCACAGCCTCTCTGGCCACCCTGGGCTGAGCACGGTGTCTGAGTCTGCTTTGACCCAGGACCA	BC??BBCB@A?BBBBB@BBBBBBCBBECCDEDDD?EC?FEHIIIIIIGIGIGBIHGDGFIIIIHGFDGEGIIIIIIHGFIGHAHGCCAGHFEDDD@@@	AS:i:0	XN:i:0	XM:i:0	XO:i:0	XG:i:0	NM:i:0	MD:Z:98	YS:i:0	YT:Z:DP

Here is the same read in bowtie1

HS2:124:C097KACXX:7:2307:12782:66337	16	4	2	255	98M	*	0	0	AGCTCCAGCCATTCCCTGGGGGCTGTCCCTGATCATCACAGCCTCTCTGGCCACCCTGGGCTGAGCACGGTGTCTGAGTCTGCTTTGACCCAGGACCA	BC??BBCB@A?BBBBB@BBBBBBCBBECCDEDDD?EC?FEHIIIIIIGIGIGBIHGDGFIIIIHGFDGEGIIIIIIHGFIGHAHGCCAGHFEDDD@@@	XA:i:0	MD:Z:98	NM:i:0

What do you guys think? Is this a SAM format issue, a bug in mpileup, something else?

[swbarnes2]

Well, for one thing, you didn't align with paired reads in bowtie 1. For another, that read in bowtie2 has a binary flag of 113, which is very strange. That's 64+32+16+1. Basically, it thinks that both of your reads are aligned in the reverse direction, which is an anomalous pair. I think mpileup by default skips anomlous pairs, so if most of your data looks like that, maybe that's why it won't call anything, because your data is so weird-loooking, it's ignoring most of it. Try running mpileup with -A, that allows it to include anomalous pairs.

Spot-check your bowtie2 .sam file, see how many reads have weird binary flags. The binary flags you want to see are 63, 99, 147, and 163. Most of your reads should have one of those four numbers. If that's not the case, your fastq's are weird.