Description

Retrieve metadata about sequencing studies performed on a subject.

The Clinical Genomics committee is reviewing Jira 35864: Add study-level metadata capabilities into the FHIR Genomics IG. It is expected that the response of this operation will be revised in the future, to align with the Jira resolution.

Test Id and test date are always returned. Additional metadata is server dependent, and specific implementations can indicate their capabilities using a FHIR Capability Statement. Frequently useful metadata includes Specimen Id and Genomic build (reference sequence assembly upon which variants were called, e.g. 'GRCh37', 'GRCh38'). Additionally, it can be useful to indicate the DNA change types or types of variants detectable by the test (e.g. 'SNV', 'InDel', 'CNV'). Servers can choose to return additional metadata.

As described in general guidance, metadata can be used to filter the results of other operations.

In some scenarios, users need to know whether or not a particular genomic region was studied by a given test (studied region), and whether or not there are uncallable subregions within regions studied (uncallable regions). Genomic sequencing tests are designed to study specific portions of a person's genome. For instance, whole genome sequencing (WGS) tests are designed to study the entire genome; whole exome sequencing (WES) tests are designed to study protein-coding exons; targeted sequencing tests are designed to only study selected regions of the genome. In many cases, while a particular region was targeted to be studied by a given test, a portion of the region might be untestable, due to patient-specific factors, reagent-specific factors, etc. Whereas labs report variants found to be present, labs cannot enumerate every variant found to be absent - when you consider that a person has 4 billion base pairs, 99% of which are the same as a reference sequence, you can imagine how big a report would need to be to enumerate every variant found to be absent! Rather, absence of a variant at a particular position can potentially be inferred by knowing the region studied, and whether or not there were uncallable subregions within the studied region.

Users can request this level of detail by including the ranges parameter, in which case each returned test will include intersecting studied and uncallable regions. Studied and uncallable regions can be voluminous - data can be curtailed by specifying appropriately narrow ranges. Range data is returned as a list of non-overlapping regions, each in zero-based RefSeq:Integer-range format (e.g. "['NC_000007.14:55173910-55174053', 'NC_000007.14:55174711-55174830', 'NC_000007.14:55181282-55181488']").