refseq_protein database is too large to run

[JiaruiChen]

I have downloaded all the genpept format from NCBI's refseq_protein database and concatenated them into one large file following the user manual and the size of the final database is 188G.

Therefore, when I use make_accession_to_taxonomy_map.pl, it always reports the error: out of memory. So I wonder whether I have followed the right step and if so, can I split the database into several sub-database and run the command for each database individually in order to avoid such an issue?

Many thanks,
James