By default, everything after an underscore or dot are disregarded at the annotation for mapping to symbols. In most cases this is "correct" as people use FEATURE_1 or FEATURE.1 for duplicated features or variants. For lipidomics this is not good because it seems underscore is an allowed character in the lipid names like "TG 14:0_16:0_16:1". Then also the other annotation columns are not correct (e.g. formula and exactmass). For lipidomics, we should retain especially when mapping to "symbols" or querying to databases, the full name. Class is luckily correct
By default, everything after an underscore or dot are disregarded at the annotation for mapping to symbols. In most cases this is "correct" as people use FEATURE_1 or FEATURE.1 for duplicated features or variants. For lipidomics this is not good because it seems underscore is an allowed character in the lipid names like "TG 14:0_16:0_16:1". Then also the other annotation columns are not correct (e.g. formula and exactmass). For lipidomics, we should retain especially when mapping to "symbols" or querying to databases, the full name. Class is luckily correct