Improve preferred names by boosting certain prefixes in choosing preferred name

[gaurav]

This PR starts the process of improving preferred names, including:

• Synonym files used to use only the label of the most preferred identifier. This PR changes that so that it uses the label of the first identifier with a label. Should close Middigital hair (OMIM:157200) does not have a preferred_label in Babel 2023may18 #142.
• A first stab at implementing "boosted prefixes" configured by Biolink type, so that we can specify that for e.g. biolink:ChemicalEntity, we should boost certain prefixes when it comes to chooses preferred_name.
  • The idea for doing this comes from reconsider name choices #158 (comment)
  • Note that this will only set the preferred_name for NameRes, not NodeNorm, which calculates it on the fly. Changing that will either require adding the preferred label to the database or to duplicate this algorithm there. I prefer the former approach, but that's definitely something we'll need to consider in the next release.

The current preferred name boost prefixes are:

  "preferred_name_boost_prefixes": {
    "biolink:ChemicalEntity": [
      "DRUGBANK",
      "GTOPDB",
      "DrugCentral",
      "CHEMBL.COMPOUND",
      "RXCUI",
      "PUBCHEM.COMPOUND",
      "CHEBI",
      "HMDB"
    ]
  }

[gaurav]

This ended up getting merged in PR #201. I'll move one minor update into PR #235 and close this PR.