Stream: genomics
Topic: Tracking WES data
Damien Jones (Jun 23 2020 at 21:56):
Hi All,
Forgive me if something like this has already been dealt with, I tried searching and didn't see anything obvious. We're trying to capture all the data generated by whole exome sequencing.
From a sample we get this data: raw reads (fastq files), an aligned genome (bam file), and variants (vcf files).
Are Observations appropriate for tracking these files? Would it make sense to bundle them in a Component?
Any guidance would be welcome.
Thanks.
Jamie Jones (Jun 24 2020 at 14:02):
Hi Damien, welcome to #genomics!
Due to the size of those files, they would likely each get their own DocumentReference resource if you wanted to put them in FHIR directly. In general, our group is focused on lab result reporting, so we have defined an extension for documents like these to be added to DiagnosticReport (see http://build.fhir.org/ig/HL7/genomics-reporting/extension-SupportingInfo.html in our Implementation Guide).
Much of the clinically actionable data represented in these files can be represented in Observation using our other structures, but we don't recommend using Observation to refer to the files directly.
You could also consider using Composition resource to group together the DocumentReferences, if it is not in the context of a ServiceRequest.
Hope this helps
Damien Jones (Jul 06 2020 at 16:58):
Thanks Jamie,
I'll check out the extension and Composition.
Bret H (Jul 27 2020 at 12:55):
look for fields that are of type 'reference.' don't know if you have read the base spec on the data type but it is worth knowing how to communicate the logical location location of a file. There is also MEDIA.
Last updated: Apr 12 2022 at 19:14 UTC
