Refactor internal data structures of seqvar query schema

[holtgrewe]

Is your feature request related to a problem? Please describe.
The seqvar query schema data structures are pretty flat. We should structure them more deeply to prevent dozens of attributes. The main offenders are CaseQuery, SequenceVariant (not part of query, but internal data structure)

Describe the solution you'd like
For CaseQuery:

• consequences..genotype - keep
• transcripts_coding, transcripts_noncoding, max_exon_dist - move to TranscriptRelated
• var_type_* - move to VariantTypeRelated
• gene_allowlist, genomic_regions - move to LocusRelated
• require_in_clinvar, clinvar_* - move to ClinvarRelated
• gnomad_*, helixmtdb_* - move to PopulationFrequencyRelated
• inhouse_* - move to InhouseFrequencyRelated

For SequenceVariant:

• chrom..ann_fields - keep
• gnomad_exomes_an..helix_het - move to PopulationFrequencies struct, rename helix_* to helixmtdb_*
• inhouse_an..inhouse_hemi - move to InhouseFrequencies struct
• call_info - keep, move up behind ann_fields

Describe alternatives you've considered
N/A

Additional context

• Dependency: Refactor seqvar query query schema and use protobuf for input #373