Hello,
I have a mostly theoretical question which my group would need to adress, though. Bismark (and possibly, other methylation mappers) have a hardcoded behavior of allowing only the unique best match to be returned as a result (Bowtie2 option -k 1, I believe).
My question would be whether this is just a recommended setting, or vitally necessary for Bismark-style methylation mapping to be correct? If so, why?
Our motivation is that we are thinking about mapping some specific data to a set of sequence subclasses (instead of the whole genome) and would like to get the 5 best hits in those sequence subclasses or so per sequence (in case more than one possible hit exists). And we're wondering whether this kind of result would be ok to interpret from Bismark methylation mapping, or why not, if not.
Many thanks
I have a mostly theoretical question which my group would need to adress, though. Bismark (and possibly, other methylation mappers) have a hardcoded behavior of allowing only the unique best match to be returned as a result (Bowtie2 option -k 1, I believe).
My question would be whether this is just a recommended setting, or vitally necessary for Bismark-style methylation mapping to be correct? If so, why?
Our motivation is that we are thinking about mapping some specific data to a set of sequence subclasses (instead of the whole genome) and would like to get the 5 best hits in those sequence subclasses or so per sequence (in case more than one possible hit exists). And we're wondering whether this kind of result would be ok to interpret from Bismark methylation mapping, or why not, if not.
Many thanks
Comment