I have a hybrid assembly (illumina + pacbio subreads - DBG2OCL assembler) for an 1.5Gb eukaryotic genome, and now I'm in doubt if I should quiver this assembly or not.
Once, as I read, quiver calls consensus based on the pacbio quality of the aligned subreads to the draft genome, what it would do to the regions that are covered only by my Illumina contigs in the hybrid assembly?
Does it make sense to run quiver in a hybrid assembly at all?
As it's a hybrid assembly, my guess is that you don't have enough coverage of pacbio data to quiver correct (<30x), otherwise it would probably make more sense to do a pacbio only assembly? If you do have ~50x pacbio data then the pacbio coverage is likely more even than the illumina, as pacbio has no GC bias, so quiver correcting will probably work. If you have contigs with no pacbio coverage, even with only a small amount of pacbio data then I wouldn't have a lot of confidence in them, given no GC bias and that it is easier to map longer reads, all contigs should have coverage. If you don't have enough pacbio data to quiver correct then I would simply use the illumina data for correction, using pilon pilon.
Tagging Dr. rhall