Add JSON schema for EVA sourced pharmacogenomics dataset

[DSuveges]

From the 23.12 release, a new dataset is being sourced by EVA: pharmacogenomics extracted from PharmGKB.

The actions:

• [x] Based on the EVA provided documentation, develop JSON schema (ideally expressed via pydantic).
• [x] Add schema to the [json-schema]() repo.
• [x] Let EVA team knows about the schema so they can use that for validation.
• [x] As this is the second dataset sourced from EVA, a new folder was needed to upload data. (gs://otar012-eva/pharmacogenomics and gs://otar012-eva/disease-target-evidence for evidence)

[DSuveges]

Based on discussion with April, the following modifications are made:

• genotypeId, variantFunctionalConsequenceId, targetFromSourceId, drugId, phenotypeFromSourceId fields are now optional.
• phenotype category become enum, with the accepted values: toxicity, efficacy, dosage, metabolism/pk, pd, other.

Once data is available, the schema will be tested and updated if required.

[DSuveges]

• Genotype IDs can have only one allele, particularly on the X chromosome, e.g. X_12885540_A_A
• I think the genotype ID regex is missing 0 for chromosomes - e.g. 20_54156202_T_C,T and 10_77596639_G_G,G both fail.
• the pgxCategory can sometimes have multiple values, e.g. efficacy;metabolism/pk... As for now, they are kept as they were. The strict enum was removed from the schema. We'll discuss this later.

[DSuveges]

• Data is uploaded to bucket.
• The developed schema successfully validates the data.