[MRG] Enable moltypes other than DNA in LCA databases

[ctb]

This PR adds support for protein, dayhoff, and hp signatures to LCA_Database and sourmash lca index. Things now work as expected on the command line, although the default scaled and ksize for sourmash lca index don't really make sense and we might want to change them in some future version.

Fixes #947

Also:

• Bumps LCA database version to 2.1.
• Also adds more/better tests for SBT index and search/gather of protein, dayhoff, and hp signatures.
• Fixes sourmash lca index doesn't give good error messages for diff ksize, non-DNA, etc #1012 poor error reporting
• Fixes sourmash signature describe misses dayhoff, hp molecule types #905 sourmash sig describe yields bad moltype
• Fixes sourmash sig describe has broken moltype output #1014 dup sourmash signature describe misses dayhoff, hp molecule types #905
• improves reporting of sourmash lca index in various ways ref classification with full length 16S gene #1000 (comment).
• fixes minor reporting bug with remnant_records in command_index, adds related lineage checking assert
• supercedes Use sourmash_args.get_moltype in "sig describe" #907

Tests to write:

• sig describe moltype stuff
• LCA_Database insert with wrong moltype
• LCA_Database insert with return val

---

• Is it mergeable?
• make test Did it pass the tests?
• make coverage Is the new code covered?
• Did it change the command-line interface? Only additions are allowed
  without a major version increment. Changing file formats also requires a
  major version number increment.
• Was a spellchecker run on the source code and documentation after
  changes were made?

[codecov]

Codecov Report

Merging #1013 into master will increase coverage by 0.07%.
The diff coverage is 98.11%.

@@            Coverage Diff             @@
##           master    #1013      +/-   ##
==========================================
+ Coverage   92.23%   92.31%   +0.07%     
==========================================
  Files          72       72              
  Lines        5397     5413      +16     
==========================================
+ Hits         4978     4997      +19     
+ Misses        419      416       -3     

Impacted Files	Coverage Δ
sourmash/lca/lca_db.py	94.62% <95.45%> (+0.28%)	⬆️
sourmash/cli/lca/index.py	100.00% <100.00%> (ø)
sourmash/lca/command_index.py	90.37% <100.00%> (+0.21%)	⬆️
sourmash/sig/__main__.py	93.37% <100.00%> (+0.66%)	⬆️
sourmash/sourmash_args.py	95.00% <100.00%> (+0.06%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 94e3244...59a6fb8. Read the comment docs.

[ctb]

@bluegenes could you review this, please? thanks!

[ctb]

Yay! Tests passed!

[bluegenes]

Other than the ksize, looks good to me!

k=19 gives 19/3=6.33, which rounds down to a k=6. It's fine for testing, but might lead to a bit of confusion if folks look to the tests to figure out how to run something (or that might just be me).

[bluegenes]

for soumash 3.x, protein kmers should be divisible by 3. Here, 19/3 = 6.3, so the actual ksize being used is k= 6.

We should maybe reimplement the check that #575/ #576 removed?

[ctb]

thanks @bluegenes! I changed the ksize=19 to ksize=57 in 20d5cf5 and 59a6fb8 - will merge if / when the tests pass!