Reading the _proteinFDR.csv output file, obtaining the stored LibMZ (reference library m/z value), then comparing it to the original has numerical errors (~1e-13). The error occurs only with peptide names containing a (UniMod:x) tag. The only time those are used together are to key a dictionary, so I'm guessing that the error stems from there.
The issue would be a problem for people reading from the _proteinFDR.csv file and relying on the values for a library lookup. It has been patched by adding small bounds on the lookup, but should be prevented from happening in the first place.
Reading the
_proteinFDR.csvoutput file, obtaining the stored LibMZ (reference library m/z value), then comparing it to the original has numerical errors (~1e-13). The error occurs only with peptide names containing a(UniMod:x)tag. The only time those are used together are to key a dictionary, so I'm guessing that the error stems from there.The issue would be a problem for people reading from the
_proteinFDR.csvfile and relying on the values for a library lookup. It has been patched by adding small bounds on the lookup, but should be prevented from happening in the first place.