View Single Post
Old 04-07-2017, 01:29 AM   #452
jweger1988
Member
 
Location: Paris, France

Join Date: Apr 2017
Posts: 37
Default

Quote:
Originally Posted by Brian Bushnell View Post
Normally, I would just map the reads with the flag "maxindel=100k" (or whatever is appropriate, given the length of the genome) and then call variants with "callvariants.sh" using the flag "rarity=0.01" to find low-frequency events. But alternatively, you can separate the reads with multiple BBMap passes, like this:

bbmap.sh in=reads.fq outm=mapped.fq maxindel=100k
bbmap.sh in=mapped.fq outm=normal.fq outu=longdeletion.fq maxindel=100k dellenfilter=20
bbmap.sh in=longdeletion.fq out=mapped_longdeletion.sam maxindel=100k
Brian, that's awesome, I didn't even know there was a variant caller in your suite. That worked well and I was able to find a bunch of deletions by using the flag "clearfilters" in callvariants.sh. However, I'm sure a lot of it is probably not real. Is there a particular parameter you suggest looking at as a measure of probability of being real? I'm not sure what some of the readouts in the output from callvariants.sh actually mean.

Thanks again.
jweger1988 is offline   Reply With Quote