in pwiz_tools/Skyline/Model/AbstractModificationMatcher.cs

                    // Some kind of garbage in peaklist, e.g fragment mass is absurdly small or large - ignore
                    // TODO(bspratt) - address Brendan's comment from pull request:
                    // "This call should be paying attention to settings and the minimum value that causes the exception reported to initiate this fix.For peptide fragment
                    // annotation, we definitely consider the settings, and since we do not rank fragments outside the instrument range. This code also strikes me as odd that you wouldn't just create the precursor
                    // and then use a precursor.ChangeSettings(Settings, diff ?? SrmSettingsDiff.ALL) to materialize all of the transitions based on the settings. That way you only write the code once to materialize
                    // transitions based on settings."
                    // In particular not ranking things outside the machine range makes sense.