I'm just starting with hh-suite, and trying to create a custom database. I thought I'd start with a single protein and see how things went from there. The instructions in the wiki indicate I should get a file (generically) called <db>_a3m.ffindex or (specifically, in my case) HEWL_a3m.ffindex.
Following the wiki (https://github.com/soedinglab/hh-suite/wiki), under "1. Creating a database of HHblits compatible MSAs - Starting with sequences" I did the following -
~/programs/static/bin/ffindex_from_fasta -s HEWL_fas.ff{data,index} HEWL_fas.fasta and got data file: HEWL_fas.ffdata index file: HEWL_fas.ffindex fasta file: HEWL_fas.fasta
Then;
~/programs/static/bin/hhblits -i HEWL_fas.fasta -d ~/uniclust/UniRef30_2020_03 -oa3m HEWL_a3m_wo_ss -n 2 -v 0
and
ls -ltr total 1484
-rw-r--r-- 1 hpowell bmm 178 Sep 22 15:51 HEWL_fas.fasta
-rw-r--r-- 1 hpowell bmm 179 Sep 22 15:54 HEWL_fas.ffdata
-rw-r--r-- 1 hpowell bmm 19 Sep 22 15:54 HEWL_fas.ffindex
-rw-r--r-- 1 hpowell bmm 678264 Sep 22 15:56 HEWL_fas.hhr
-rw-r--r-- 1 hpowell bmm 364359 Sep 22 15:56 HEWL_a3m_wo_ss
The wiki says - To be sure that everything went smoothly, check that the number of lines in <db>_a3m.ffindex is the same as the number of lines in <db>_fas.ffindex.
Pretty obviously I'm doing something wrong (because an expected output file (HEWL_a3m.ffindex) is not created), so a nudge in the right direction would be appreciated!
Harry
Thanks for the rapid answer - from what you say (which makes good sense and is pretty well what I would have expected), the Wiki is incorrect because the order of commands I used is the order given in the example in the Wiki.
I'll ask the authors for advice...