Filter transcripts

[ahc87]

https://www.ebi.ac.uk/panda/jira/browse/EA-1037

This change adds the option to filter the transcripts returned by transcripts_page. Here is an example query:

query {
  gene(by_id: { 
    genome_id: "a7335667-93e7-11ec-a39d-005056b38ce3", 
    stable_id: "ENSG00000109339" 
  }) {
    transcripts_page(page: 1, per_page:10, 
    filters: {
      biotype: ["protein_coding", "nonsense_mediated_decay"],
      tsl: ["tsl1", "tsl2", "tsl3"]
    }
  ) {
    transcripts {
      stable_id
      unversioned_stable_id
      so_term
      metadata {
        tsl {
          value
        }
        biotype {
          value
        }
      }
    }
    page_metadata {
      page
      per_page
      total_count
    }
  }
}
}

[ahc87]

We decided on how we should do filter groups during the Thoas prioritisation meeting, documented here: https://www.ebi.ac.uk/seqdb/confluence/display/EA/Meeting+-+2023-01-10. This PR hardcodes metadata types in the GraphQL schema, which is incompatible with the agreed solution for filter groups. Therefore this PR needs to be updated.