Bowtie option --best --strata for unique mapper in the genome

[StephaniePi83]

Hi,

I have some trouble with alignment using bowtie.
I'm searching for reads that map at a unique place on the genome, so, i take 1 reads that i know it's a unique mapper and it map the genome 100%, and i try to align it with bowtie with the following command, allowing up to 2 mismatches (because for the other reads i want to allow up to 2 mismatches).

For verification :

[B]bowtie -v 0 -m 1 -t droso_build -c ATGCGAGAGTGGT[/B]
Time loading forward index: 00:00:01
Time for 0-mismatch search: 00:00:00
0	+	chr2L	5308	ATGCGAGAGTGGT	IIIIIIIIIIIII	0	
# reads processed: 1
# reads with at least one reported alignment: 1 (100.00%)
# reads that failed to align: 0 (0.00%)
Reported 1 alignments to 1 output stream(s)
Time searching: 00:00:01
Overall time: 00:00:01

But, when i try to align the other reads (allowing up to 2mismatches), this reads have no hit :

[B]bowtie -m 1  -v 2 -t droso_build -c ATGCGAGAGTGGT[/B]
Time loading forward index: 00:00:00
Time loading mirror index: 00:00:01
End-to-end 2/3-mismatch full-index search: 00:00:00
# reads processed: 1
# reads with at least one reported alignment: 0 (0.00%)
# reads that failed to align: 0 (0.00%)
# reads with alignments suppressed due to -m: 1 (100.00%)
No alignments
Time searching: 00:00:01
Overall time: 00:00:01

But,when i do:

And whan i add -a --best --strata i get :

bowtie -a -v 2 -m 1 --best --strata -t droso_build -c ATGCGAGAGTGGT
Time loading forward index: 00:00:00
Time loading mirror index: 00:00:01
End-to-end 2/3-mismatch full-index search: 00:00:00
0	+	chr2L	5308	ATGCGAGAGTGGT	IIIIIIIIIIIII	0	
# reads processed: 1
# reads with at least one reported alignment: 1 (100.00%)
# reads that failed to align: 0 (0.00%)
Reported 1 alignments to 1 output stream(s)
Time searching: 00:00:01
Overall time: 00:00:01

I'm really lost, is the -a --best --strata a good solution ?

Thanks in advance for your reply.

[vsravi]

Reporting uniquely mapped reads with bowtie

I have a similar problem - I want bowtie to report only uniquely mapped reads. I figured using the flags -n 3 --best -m1 will accomplish this:
--best should ensure that only the best alignments among the hits are considered, and -m 1 should ensure that the alignment is reported only when there is a unique best alignment.

But, it doesn't seem to work that way.
bowtie -n 3 -a returns two alignments for the read:

alignment1 CIGAR: 48M flags: XA:i:0	MD:Z:48	NM:i:0
alignment2 CIGAR: 48M flags: XA:i:0	MD:Z:0C47	NM:i:1

Clearly, the second alignment is not the same as the first.
bowtie -n 3 --best returns only the first alignment:

alignment1 CIGAR: 48M flags: XA:i:0	MD:Z:48	NM:i:0

Everything looks good so far. Now, I add the -m 1 flag, to report only unique reads, but:
bowtie -n 3 --best -m 1 reports no alignments for this read:

# reads processed: 1
# reads with at least one reported alignment: 0 (0.00%)
# reads that failed to align: 0 (0.00%)
# reads with alignments suppressed due to -m: 1 (100.00%)
No alignments

Am I doing anything wrong? Any help is appreciated.