--dpad --gbar

[zznx]

Hi, I'm reading the manual
I don't know what's going on here, but if there's a problem, how can I solve it myself? Can you explain? Or provide relevant literature, thank you

https://bowtie-bio.sourceforge.net/bowtie2/manual.shtml#bowtie2-options-align-paired-reads
--dpad <int> | "Pads" dynamic programming problems by <int> columns on either side to allow gaps. Default: 15.
-- | --
--gbar <int> | Disallow gaps within <int> positions of the beginning or end of the read. Default: 4.

[ch4rr0]

Hello,

--gbar <int> | Disallow gaps within <int> positions of the beginning or end of the read. Default: 4.

Given the following reference:

>a
GGAATATTTGCGATTTGCCATTTTCTCTCAAGAGT

and read:

>r1
GGAATAGCGATTTGCCATTTTCTCTCAAGAGT

Trying to align the read to reference will cause bowtie2 to report a gap in the alignment i.e.
in order to transform the reference to the read the 3 Ts, starting at position 7, would have
to be deleted from the reference (you can see this represented in the CIGAR as well as the MD:Z).

GGAATATTTGCGATTTGCCATTTTCTCTCAAGAGT
||||||   ||||||||||||||||||||||||||
GGAATA---GCGATTTGCCATTTTCTCTCAAGAGT
12345678901234567890123456789012345

./bowtie2-align-s -x /tmp/out -f read.fq  -a --overhang --gbar 4
r1	0	a	1	255	6M3D26M	*	0	0	GGAATAGCGATTTGCCATTTTCTCTCAAGAGT	IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII	AS:i:-14	XN:i:0	XM:i:0	XO:i:1	XG:i:3	NM:i:3	MD:Z:6^TTT26	YT:Z:UU

This gap happens within 6 bases of the start of the read so bowtie2 will report this alignment
as valid since it does not violate the default for --gbar which is 4.

If we change --gbar to 6 then bowtie2 will no longer consider this alignment as valid.

./bowtie2-align-s -x /tmp/out -f read.fq  -a --overhang --gbar 6
1 reads; of these:
  1 (100.00%) were unpaired; of these:
    1 (100.00%) aligned 0 times
    0 (0.00%) aligned exactly 1 time
    0 (0.00%) aligned >1 times
0.00% overall alignment rate
@HD	VN:1.5	SO:unsorted	GO:query
@SQ	SN:a	LN:35
@PG	ID:bowtie2	PN:bowtie2	VN:2.5.4	CL:"/home/rcharles/src/git/bowtie2/bowtie2-align-s -x /tmp/out -f read.fq -a --overhang --gbar 6"
r1	4	*	0	0	*	*	0	0	GGAATAGCGATTTGCCATTTTCTCTCAAGAGT	IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII	YT:Z:UU

I will need to get back to you with an example explaining --dpad. In the meantime I hope this helps.

[zznx]

Yes, I know what you mean, it's a kind of assumption that I think there should be no gap in the first four bases, or it's a rule summarized by a lot of data. Which one is it?