View Single Post
Old 10-02-2017, 04:03 AM   #36
GenoMax
Senior Member
 
Location: East Coast USA

Join Date: Feb 2008
Posts: 6,780
Default

@Wouter: @Brian would be along with an official answer but this may be because most NGS aligners are non-determinstic by default i.e. they may not produce exactly the same answer for each run.

bbmap.sh provides a "determinisitc" flag that you can set to "true" to get exactly identical results each time. I don't see bbsplit listed as having that flag but it may support it since it is based on bbmap. @Brian will confirm.

You would also need to consider how similar your references are and how you are choosing to handle multi-mappers (reads which map across genomes). Are you leaving the ambiguous2= option default. It may help to post the entire bbsplit command you are using for this analysis.
GenoMax is offline   Reply With Quote