Record hmm_mincov params in startup message

v1.10.2

Rename to 1.10.2

Separate hmm_mincov for proteins and domains

Separate hmm_mincov for proteins and domains

Pin and clean up dependencies

Dependencies were unnecessarily wide and without versions -- fixed.

dplyr instead of data.table

Reverted to dplyr from data.table for 10x better performance (in tests).

Handle empty files

Empty tblout and domtblout files caused errors -- fixed.

Revert .groups

.groups in summarize caused creation of an extra field in some cases. Probably due to earlier dplyr versions, but can't confirm that. Deleted this parameter, which means there are now warnings.

Don't use this, it was just because dplyr < 1.0.0. was used.

Switch to gawk

Standard awk had trouble with the fasta reading, switching to gawk.

Correct domains file

Fixed a long standing mistake in how the domains table was created. Now it's exactly like the proteins table but contains hits to profiles with prank == 'domain'.

Better logging

• Write missing genome accessions to file.
• Better informational logging.

More GTDB genome information

Many extra fields from the GTDB metadata file is now included in the feather output from pf-classify.r.