This should feed to one.site.per.read in fit-hhm.R
I recommend that the default be FALSE.
Peter: My reason for arguing the default be False is that the parameter is actually inaccurately named. If TRUE, the code does not take one site per read, it takes one site per set of contiguous base pairs. Thus, for example, if someone tried to use light shotgun reads, then long contiguous regions would result in one site taken in each contig. We often see reads overlapping with WMD-MSG, so much of these data are discarded with TRUE. This will become increasingly problematic as read depth increases with Illumina and other sequencing technologies.
Add parameter to msg.cfg file
one.site.per.contig True or False
This should feed to one.site.per.read in fit-hhm.R
I recommend that the default be FALSE.
Peter: My reason for arguing the default be False is that the parameter is actually inaccurately named. If TRUE, the code does not take one site per read, it takes one site per set of contiguous base pairs. Thus, for example, if someone tried to use light shotgun reads, then long contiguous regions would result in one site taken in each contig. We often see reads overlapping with WMD-MSG, so much of these data are discarded with TRUE. This will become increasingly problematic as read depth increases with Illumina and other sequencing technologies.