bowtie2 fragment length

[MerFer]

Dear Bowtie2 user,

I have a question regarding the fragment length issue.

The reads that mapped to the different chromosomes sometimes have a fragment length of 0 but sometimes non-0. From the bowtie2 manuel
I understand that the fragment length has to be non-0 if the mates mapped to the same chromosome.

"Inferred fragment length. Size is negative if the mate's alignment occurs upstream of this alignment. Size is 0 if the mates did not align concordantly. However, size is non-0 if the mates aligned discordantly to the same chromosome."

Here is an example from my bowtie2 output:
Example for 0 fragment length

MISEQ:52:000000000-A4DCH:1:1101:16444:3647 65 chr9 71742006 24 22M = 3007124 0 TTAATGTCTGTCTCCCTTACTG FFBFFFFGGGGFGGGGGHHHHG AS:i:-5 XN:i:0 XM:i:1 XO:i:0 XG:i:0 NM:i:1 MD:Z:21A0 YT:Z:UP
MISEQ:52:000000000-A4DCH:1:1101:16444:3647 129 chr9 3007124 0 29M = 71742006 0 TCGTCATTTTTCAAGTCGTCAAGGGGATG ABBBBBFFFFFFGGGGGGGGGGCHFEEGG AS:i:-5 XN:i:0 XM:i:1 XO:i:0 XG:i:0 NM:i:1 MD:Z:23T5 YT:Z:UP

Example for non-0 fragment length

MISEQ:52:000000000-A4DCH:1:1101:17281:1944 65 chr14 54510108 42 23M = 52820128 -1690003 TTAAAGGTAAAAACCACCAACAT BFAFFFF1BGGGGFEF1AGFEGH AS:i:0 XN:i:0 XM:i:0 XO:i:0 XG:i:0 NM:i:0 MD:Z:23 YS:i:0 YT:Z: DP
MISEQ:52:000000000-A4DCH:1:1101:17281:1944 129 chr14 52820128 42 31M = 54510108 1690003 CCTTGGAAGCCCACTGGAAAAAATGACAGAT AAABAFFFFCFBGGGGGGGGGGGHHHHGHFG AS:i:0 XN:i:0 XM:i:0 XO:i:0 XG:i:0 NM:i:0 MD:Z:31 YS:i:0 YT:Z: DP

Do I miss something or is it a bowtie2 bug?

Thanks in advance,

Mary

[westerman]

Well both examples are discordant within the same chromosome (chr9 for the first example, chr14 for the second). However in your first example the MAPQ of the second read is 0. This may cause bowtie2 to not want to infer a TLEN (fragment size) thus setting it to 0.