bowtie -M

[masylichu]

I wanna map the small RNA reads to the reference C. elegans genome.

for example:
>test.fa
GGGAATCCTGGATGTTGTAAGCTT

which has 15 copies in the genome.


BLAT Search Results

ACTIONS QUERY SCORE START END QSIZE IDENTITY CHRO STRAND START END SPAN
---------------------------------------------------------------------------------------------------
browser details YourSeq 24 1 24 24 100.0% V - 17115871 17115894 24
browser details YourSeq 24 1 24 24 100.0% V - 17117831 17117854 24
browser details YourSeq 24 1 24 24 100.0% V - 17118805 17118828 24
browser details YourSeq 24 1 24 24 100.0% V + 17122985 17123008 24
browser details YourSeq 24 1 24 24 100.0% V + 17123961 17123984 24
browser details YourSeq 24 1 24 24 100.0% V + 17124937 17124960 24
browser details YourSeq 24 1 24 24 100.0% V + 17125913 17125936 24
browser details YourSeq 24 1 24 24 100.0% V + 17128469 17128492 24
browser details YourSeq 24 1 24 24 100.0% V + 17129441 17129464 24
browser details YourSeq 24 1 24 24 100.0% V + 17130427 17130450 24
browser details YourSeq 24 1 24 24 100.0% V + 17131403 17131426 24
browser details YourSeq 24 1 24 24 100.0% V + 17428943 17428966 24
browser details YourSeq 24 1 24 24 100.0% V + 17122009 17122032 24
browser details YourSeq 24 1 24 24 100.0% V + 17121033 17121056 24
browser details YourSeq 24 1 24 24 100.0% V + 17120058 17120081 24

i use the command line:

bowtie -t -v 0 ~/WormBase190/WS190/genome.ws190 -M 25 -f test.fa test_map

M option means that but reports 1 random hit if no more than <int>.

However, every times, i run the command, it always reports
test - V 17118804 AAGCTTACAACATCCAGGATTCCC IIIIIIIIIIIIIIIIIIIIIIII 2

Why, the M options doesnot report random-selected one hit ?

[mgogol]

It's not possible you're using bowtie 2, is it?

[sdriscoll]

if you switch to -m and also use -a do you get all of the same alignment locations that BLAST reported?

[sdriscoll]

Wait...I've got it. Try using -M 1 instead of -M 25. The bowtie manual says that if the read can be aligned more than <int> times then one alignment can be reported at random. You say your sequence can align in 15 locations but you used -M 25 meaning the limit has not been reached to where it will report the single random alignment for that read. So if you had also specified -k 25 then you would probably see all 15 alignments of your sequence.

[masylichu]

YES. so -a options should not be used under this condition.

[masylichu]

The bowite manual says "-M <int> like -m, but reports 1 random hit (MAPQ=0); requires --best "

This is to say if the read align more than <int>, then it will suppress all the alignments.

As what you said, -M 1 here will reports no alignment.

Wait...I've got it. Try using -M 1 instead of -M 25. The bowtie manual says that if the read can be aligned more than <int> times then one alignment can be reported at random. You say your sequence can align in 15 locations but you used -M 25 meaning the limit has not been reached to where it will report the single random alignment for that read. So if you had also specified -k 25 then you would probably see all 15 alignments of your sequence.

[masylichu]

It's not possible you're using bowtie 2, is it?

The bowtie2 is particularly good at aligning reads of about 50 up to 100s or 1,000s of characters. Here, i used bowtie to map the small reads.

[mgogol]

I just asked, because the -M in bowtie 2 has a different meaning. I don't really know what's going on, sorry. You could try bwa?

[mgogol]

Oh, I just read this, I guess you need to set the seed differently to get a different read.

From the manual:
"This randomness flows from a simple seeded pseudo-random number generator and is deterministic in the sense that Bowtie will always produce the same results for the same read when run with the same initial "seed" value (see --seed option)."

http://bowtie-bio.sourceforge.net/manual.shtml#bowtie-options-seed