This is a re-analysis of the dataset deposited in PXD002171 (DOI: 10.1021/acs.jproteome.5b00420, PubMed: 26626371) with databases created by ProtGraph. We only changed the data processing, the sample creation and processing is the same as in the original submission. ProtGraph allows to incorporate variant and further feature information, which changes the amino acid composition of consensus/canonical sequences in a database and is commonly not used in the database analysis.