Add a Crimean-Congo hemorrhagic fever virus dataset

[anna-parker]

This PR updates the CCHF dataset started in #200

It does the following:

1. adds basic QC
2. updates alignment parameters:

• Decreases the minSeedCover to 10% - this means the total number of sequences on NCBI virus that cannot be aligned using this dataset drops from 148 to 48.
• Decreases the kmer size to 7 for segment M and S - further drop to only 34 sequences.
• Add retryReverseComplement - able to align 25 further sequences (drop to only 9)
• Increases the the gapOpenScores to 18, 19 and 20 for the S and L segment and to 24, 26 and 28 for the M segment (see reasoning below)

3. Adds clades for the S segment using the annotation defined in
   Serena A. Carroll, Brian H. Bird, Pierre E. Rollin, Stuart T. Nichol,
   Ancient common ancestry of Crimean-Congo hemorrhagic fever virus (2010), link. The nextstrain build used for this can be found here: https://github.com/neherlab/CCHFV

More Insertions or Mutations?

I see high divergence right before the terminals of all segments, especially the M segment, this is partially in line with previous findings: "Prominent features of the M RNA segment are a high degree of divergence at the first part of the M genome, along with conservation of the middle and last regions and the 10-nt termini, which are conserved in all nairoviruses" (https://pmc.ncbi.nlm.nih.gov/articles/PMC2730268/).

After discussion with @corneliusroemer I increased the gapOpen score in all segments, but especially in the M segment, to make insertions more expensive. For most sequences the number of insertions relative to the parent is now comparable to the number of mutations relative to the parent, whereas before it was approx. double that.

M segment alignment, only sequences with high coverage, default gapOpen score:

Current M segment alignment:

Preview

https://clades.nextstrain.org/?dataset-server=https://raw.githubusercontent.com/anna-parker/nextclade_data/cchfv_updates/data_output