Hi, some gene and inhibitor searching I have been doing is confounded by what I can term MeSH "back-mappings".

Specifically; "BACE1 protein, human"[All Fields] brings back 1466 entries, but the trouble is they go back to 1991.

This is 8 years before BACE1 was discovered and named (by ourselves PMID:10656250 and others) and MeSH only introduced the term "BACE1" in 2004.

I can see the utility of this in some cases where old synonyms can be mapped to new synonyms but for this protein name the backwards transitivity (via beta-secretase as a phenomenological description in pre-BACE1 papers) is perilous and produces spurious recall at default settings.

I know I can use the advanced search filters to circumvent this (and this is certainly what the NCBI help desk will say) but I was wondering if anyone (from NLM/MeSH even?) had background information on this as a general problem.