As I checked in
the mrsFAST alignment paper in http://www.nature.com/nmeth/journal/v7/n8/full/nmeth0810-576.html,
when BWA is set with the parameter N (get all the
mappings), BWA reports "(1) many repetitions (the same locus
reported multiple times - sometimes due to several alignment scripts with the
same number of errors), (2) sequences involving the symbol N (i.e. aNy nucleic
acid) -BWA assumes that N correctly maps to any symbol from the four letter DNA
alphabet - these, which constitute about %10 of the mapping locations returned
by BWA are avoided by mrsFAST."
If this is true,
is there a way to avoid at least the first "bug"?