So you object to my removals of a broad "molecular biology" statement as main subject of an item about an article.
Generally, as you know, I concentrate on MeSH starred terms, and I work now to give the exact MeSH term, not something broader. It seems that quite broad terms are not really considered to be so bad, but I think the main subject statements should be referenced. And where there is a reference, the statements should be made accurate according to that reference. You and at least one other editor add comparable statements from other such good sources.
There are also main subject statements coming from the author-supplied keywords. Those index terms are strings, and need to be matched to Wikidata items. I add those, with reference and two qualifiers so that everyone can see what is happening.
Finally, there are inferred statements. There are many added by the Source MD tool for item creation, and that tool adds many broad statements, and many bad statements also. I have a page for tracking seriously bad statements and a high proportion of the errors on that page seem to come from that tool.
From the early days one can see many broad statements such as "organic chemistry" and "catalysis" that might be true, if there an enzyme involved, but really aren't helpful. There are a large number of statements "inferred from title" that are unhelpful in various ways, because automated text-mining of titles is not a great technique.
So, if you want me not to delete a main subject like "molecular biology" which is an inference of a certain kind, you should explain how the inference is made, and why it is helpful. Because much of what happens seems to me not to be helpful. Charles Matthews (talk) 12:42, 21 November 2022 (UTC)