Implement a SequenceTyper

[apetkau]

Since I have access to mutations/mlst/kmers (sketches) I can take advantage of all of these to do sequence typing.

So, one idea is to implement a SequenceTyper class which gets passed to my query API. For example, say I wanted to type sequences based on mutations (e.g., https://github.com/cov-lineages/constellations of SARS-CoV-2). I could create a class like so:

class SARSCov2ConsellationsTyper(SequenceTyper):

    def type_query(sequence_type: str, query: SamplesQuery) -> SamplesQuery:
        samples_with_mutations = query
        for mutation in get_mutations_for_sequence_type(sequence_type):
            samples_with_mutations = samples_with_mutations.hasa(mutation)
            # Actual implementation requires counting mutations to account for missing data
        return samples_with_mutations

Where type_query() is a method to implement to select samples of a particular type.

This could then be used as:

sarscov2_typer = SARSCov2ConstellationsTyper.create(input_data)
delta_samples = query.isa('Delta', kind=sarscov2_typer)
# delta_samples now contains a selection of samples which are of type "Delta"

This wouldn't be limited to mutations/SARS-CoV-2 though. I could do the same thing with MLST results, with kmer sketches, with phylogenetic trees, with attached metadata, or a combination of all of them.