error with zygosity/dna change type for Phenopacket input

[aslgraefe]

When importing a sample RareLink-CDM Phenopacket, an error is shown for the way we have zygosity and the dna change type encoded:

        "genomicInterpretations": [
          {
            "subjectOrBiosampleId": "1",
            "variantInterpretation": {
              "variationDescriptor": {
                "id": "829137227d3443f2aa73fc94a4f639",
                "geneContext": {
                  "valueId": "HGNC:6932",
                  "symbol": "MC4R"
                },
                "expressions": [
                  {
                    "syntax": "hgvs",
                    "value": "NM_005912.3:c.751A>C"
                  }
                ],
                "structuralType": {
                  "id": "LOINC:LA6694-9",
                  "label": "Frameshift"
                },
                "allelicState": {
                  "id": "LOINC:LA6706-1",
                  "label": "Heterozygous"
                }

Error for zygosity:

Cell In[5], [line 5](vscode-notebook-cell:?execution_count=5&line=5)
      [3](vscode-notebook-cell:?execution_count=5&line=3) cohort_creator = configure_caching_cohort_creator(hpo)
      [4](vscode-notebook-cell:?execution_count=5&line=4) pp_dir = '/Users/adam/Documents/git/gpsea-cs/irud_test/phenopackets'
----> [5](vscode-notebook-cell:?execution_count=5&line=5) cohort, qc_results = load_phenopacket_folder(pp_dir, cohort_creator)

File ~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:440, in load_phenopacket_folder(pp_directory, cohort_creator, validation_policy)
    [437](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:437) pp_files = _find_phenopacket_files(pp_directory)
    [439](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:439) # Map to patients
--> [440](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:440) return load_phenopacket_files(
    [441](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:441)     pp_files=pp_files,
    [442](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:442)     cohort_creator=cohort_creator,
    [443](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:443)     validation_policy=validation_policy,
    [444](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:444) )

File ~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:464, in load_phenopacket_files(pp_files, cohort_creator, validation_policy)
    [447](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:447) def load_phenopacket_files(
    [448](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:448)     pp_files: typing.Iterator[str],
    [449](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:449)     cohort_creator: CohortCreator[Phenopacket],
    [450](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:450)     validation_policy: typing.Literal["permissive", "lenient", "strict"] = "permissive",
    [451](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:451) ) -> typing.Tuple[Cohort, PreprocessingValidationResult]:
    [452](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:452)     """
    [453](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:453)     Load phenopacket JSON files, validate the data, and assemble into a :class:`~gpsea.model.Cohort`.
    [454](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:454) 
...
     [73](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_phenopacket.py:73)     return Genotype.HEMIZYGOUS
     [74](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_phenopacket.py:74) else:
---> [75](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_phenopacket.py:75)     raise ValueError(f"Unknown genotype {genotype}")

ValueError: Unknown genotype Heterozygous

-> this is fixed if I simply change the "Heterozygous" to a small capitalised "heterozygous". However, the LOINC label was automatically fetched from Bioportal. Given the Phenopackets also allow LOINC, could its label also be allowed to work?

Error for DNA Change type:

ValueError                                Traceback (most recent call last)
Cell In[6], [line 5](vscode-notebook-cell:?execution_count=6&line=5)
      [3](vscode-notebook-cell:?execution_count=6&line=3) cohort_creator = configure_caching_cohort_creator(hpo)
      [4](vscode-notebook-cell:?execution_count=6&line=4) pp_dir = '/Users/adam/Documents/git/gpsea-cs/irud_test/phenopackets'
----> [5](vscode-notebook-cell:?execution_count=6&line=5) cohort, qc_results = load_phenopacket_folder(pp_dir, cohort_creator)

File ~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:440, in load_phenopacket_folder(pp_directory, cohort_creator, validation_policy)
    [437](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:437) pp_files = _find_phenopacket_files(pp_directory)
    [439](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:439) # Map to patients
--> [440](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:440) return load_phenopacket_files(
    [441](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:441)     pp_files=pp_files,
    [442](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:442)     cohort_creator=cohort_creator,
    [443](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:443)     validation_policy=validation_policy,
    [444](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:444) )

File ~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:464, in load_phenopacket_files(pp_files, cohort_creator, validation_policy)
    [447](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:447) def load_phenopacket_files(
    [448](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:448)     pp_files: typing.Iterator[str],
    [449](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:449)     cohort_creator: CohortCreator[Phenopacket],
    [450](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:450)     validation_policy: typing.Literal["permissive", "lenient", "strict"] = "permissive",
    [451](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:451) ) -> typing.Tuple[Cohort, PreprocessingValidationResult]:
    [452](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:452)     """
    [453](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:453)     Load phenopacket JSON files, validate the data, and assemble into a :class:`~gpsea.model.Cohort`.
    [454](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_config.py:454) 
...
    [819](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_phenopacket.py:819)         return None
    [820](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_phenopacket.py:820) else:
--> [821](https://file+.vscode-resource.vscode-cdn.net/Users/adam/Documents/git/gpsea-cs/irud_test/~/Documents/git/gpsea-cs/gpcvenv/lib/python3.12/site-packages/gpsea/preprocessing/_phenopacket.py:821)     raise ValueError(f"Unknown structural type {structural_type}")

ValueError: Unknown structural type LOINC:LA6694-9

Same here, why are LOINC values not allowed, although possible in Phenopackets themselves?

[pnrobinson]

It is not a good idea to use LOINC for genetics/genomics - none of the downstream software knows what to do with it.
Please use sequence ontology for the allelic state (see examples in phenopacket store).
Also, "frameshift" refers to the consequence of a (usually small) variant, it is not a structural type, so this is wrong

 "structuralType": {
                  "id": "LOINC:LA6694-9",
                  "label": "Frameshift"
                },

You can/should just leave this out for HGVS-encoded variants (downstream software can figure out what the consequence is). We also use sequence ontology for structural variants such as translocation.

[aslgraefe]

Thank you!
That is interesting to hear as the official HL7 FHIR Genomics Reporting recommends LOINC for its allelic state: https://hl7.org/fhir/uv/genomics-reporting/STU3/StructureDefinition-variant-definitions.html#diff_Observation.component:allelic-state.value[x] (which is why we have it in the rarelink-cdm)... I will incorporate an automatic mapping to the sequence ontology for these. This is kind worrysome for people learning all of this and HL7 recommends these..

When deleting them I get this errors when running qc_results.summarize:

Validated under permissive policy
Showing errors and warnings
Phenopackets
  patient #0
   errors:
   ·Individual 1 has no genotype data (variants) to work with. Add variants or remove the individual from the analysis
    variants
      #0
       warnings:
       ·Missing required `structural_type` field
       ·Could not extract the information for large SV annotation. Remove variant from testing
  patient #1

for a variant like this

  "interpretations": [
    {
      "id": "test_1-interpretation-0",
      "progressStatus": "SOLVED",
      "diagnosis": {
        "disease": {
          "id": "OMIM:618406",
          "label": "BODY MASS INDEX QUANTITATIVE TRAIT LOCUS 20"
        },
        "genomicInterpretations": [
          {
            "subjectOrBiosampleId": "1",
            "interpretationStatus": "CAUSATIVE",
            "variantInterpretation": {
              "variationDescriptor": {
                "id": "ecdb718df3894ab1a99226d7f2b784",
                "geneContext": {
                  "valueId": "HGNC:6932",
                  "symbol": "MC4R"
                },
                "expressions": [
                  {
                    "syntax": "hgvs",
                    "value": "NM_005912.3:c.751A>C"
                  }
                ],
                "allelicState": {
                  "id": "GENO:0000135",
                  "label": "heterozygous"
                }

[ielis]

Hello @aslgraefe

I think that you need to change "syntax": "hgvs" to "syntax": "hgvs.c" to make this work.

I do not think we describe this in the docs, though.. 😢

[pnrobinson]

For FHIR it is OK; but this needs to be translated into a genomics-friendly ontology for downstream (IMHO).
Also, the above variant is NOT a structural variant, it is a single nucleotide vairant (SNV) and so it does not have a structural type. The QC script is incorrect in demanding this ·Missing required structural_type field
·Could not extract the information for large SV annotation. Remove variant from testing
patient #1

[aslgraefe]

great, thank you! Now it works.

I will fix these things in rarelink. To summarise what we could improve in the phenopackets docs:

• Variant.allelic_state: note that LOINC is not good for analysis, transformation required to GENO or SO
• syntax hgvs does not suffice, but hgvs.c or hgvs.g better for analysis
• the structural_type stuff probably resolves when taking real variants from individuals - this was just a random variant chosen. but still perhaps good to note something about this. But I personally do not fully understand this yet, I'll need to read into some genetics text books (med school does not prepare for any of this reallly...)