While the sequenced type strains for each species were used, there is still potential for polymorphic differences between the cultured strains and the reference genome. Classification and consequent masking of these genomic locations in the reference allows more accurate modeling of error rates in downstream analyses.
