How can I tell hmmscan to take the top scoring target and output that instead of querying each contig node against the hmm profile database?

I created a hmm profile using the 735 BacMet v2 experimentally confirmed protein database using jackhmmer against the 151k protein homologs. I built a hmm profile for each protein and then concatenated them, then hmmpress'd them.

See here: C: Using blast or similar tool to annotate ORFs using Bacmet database

I used this file to run a hmmscan against some ORFs I predicted using prodigal but when I run it, it tries to search each node against all the homologs. I tried the --cut_ga option but then it returns no hits.

Please help.

Thanks!