search

antigenomics / vdjdb-db

🗂️ [vdjdb.cdr3.net is up and running] Git-based TCR database storage & management. Submissions welcome!
https://vdjdb.cdr3.net
Other
131 stars 27 forks source link

`antigen.gene` / `antigen.species` mix up for some human epitopes #368

Open JamieHeather opened 4 months ago

JamieHeather commented 4 months ago

I've just updated an analysis to use the May 2024 release and noticed that there's a few entries for anti-human TCRs that appear to have the antigen.gene and antigen.species fields swapped around. There's just under 90 of them, from a handful of different sources - see switched-antigen-gene-species.txt.

(In double checking this I also noticed that sometimes a human protein name is used instead of the gene name in the antigen.gene field, e.g. p53 instead of TP53 or NY-ESO-1 instead of CTAG1B. I'm not sure if it matters, but I guess it's worth maybe flagging for people who want to cross-reference between datasets by gene symbols.)

mikessh commented 4 months ago

Yep, just spotted that too, sorry - got a bit distracted with Docker issues) It matters as some "self-antigens" become missing. Thanks for the fix!