View Single Post
Old 07-24-2014, 04:59 AM   #3
shambo
Junior Member
 
Location: London

Join Date: May 2012
Posts: 6
Default

You're absolutely right, there is a 3 base PAM (NGG) on the end of the sequence. In my case all the sequences had the NGG when they were supplied to the mapper.

You are also right when you say the R package takes the preservation of the NGG into account when looking for off-target sites, which is why I was surprised when bwa returned fewer when it is free to mis-match anywhere within the sequence. Mu thought was to use the mapper and then screen for hits where the NGG rule is preserved.

I was pretty much copying what was written here

http://www.sanger.ac.uk/htgt/wge/crispr_help

so in principle it seems it should work, so why it doesn't is bit of a mystery to me right now.

Thanks,

Sham
shambo is offline   Reply With Quote