Protein accession to NCBI-taxonomy mapping file update


I have just uploaded taxonomy mapping files dating March 2018


Thank you Daniel. I re-analyzed my data using the new mapping file, but found that many viral reads that had previously been assigned using the May 2017 mapping file were unassigned using the March 2018 mapping file. Others that had been more specifically classified were simply denoted as viruses after the mapping file update. When I looked at the reads and individually checked them via blastx, they were given the same assignments as those analyzed using the May 2017 mapping file. I noticed a note on the download page that says if there is no class classification, then the read is assigned the majority class of equivalent classifications. Since many viral entries are not given class annotations, this may be part of the reason I am losing so many reads with the undated mapping file. For this reason, I am planning to stick with the May 2017 mapping file, and only update the nr database for use with DIAMOND. Does this sound reasonable?


Are you sure that you used exactly the same settings to reanalyze the data? Could you send me a small file of virus reads that exhibit the problem, then I’d take a look to see what is going on


Daniel, I have sent you an email at Thank you!


that probably won’t reach me… please send to my university email address


OK I’ve sent it to your university email address.


Hi Daniel,

Will it be possible to update the protein accession mapping file?



will do so within the next week or two


Just updated taxonomic mapping and also InterPro2GO mapping


Dear Daniel.

I tried to download the mapping file but failed.
When I try open this link (, show: “We’re sorry, but page does not exist”.

I even tried last version of M. Again no.



That appears to have been a temporary outage, the webpage is accessible again