5 Conclusions
We have shown that proteogenomic analysis is beneficial in the
assessment of the virulence of P. larvae . The concept that we
present is that diversity/variations in successful identifications from
draft genome assemblies may be a factor underlying differences among
strains. However, such an analysis is not appropriate for all types of
samples. The method is useful in that proteogenomic search is performed
on a high-quality MS/MS dataset of the exoprotein fraction of different
type strains. That dataset, although composed of different genotypes,
serves in the first step as one source of data, and interstrain
differences in the proteome are not considered; however, the success of
the decoy database components is considered. In the next step, the
differences in identifications are used for the identification of
expression differences among the type strains. Finally, the unique
differences in the identification success of databases and protein
markers that were mined from proteome differences are connected. This
study provided novel markers that could be further studied. Our findings
support the idea that many differently virulent strains can exist and
may not be limited to the current ERIC I–V genotypes. Various P.
larvae strains have specificities that make them unique and different
from ERICs. Finally, we stress the importance of sharing high-quality
proteomics data, making them available for multiple reanalyses, since we
also reused previously reported MS/MS data. Secondary data evaluation
using an alternative approach and/or updated databases may provide
alternative interpretations and identify novel markers.