Post v1.0.0 new features

[e-t-k]

This discussion is to track new features in the interface design that we would like to add post v1.0.0 release. This can be large features that will be added in v2.0.0, or smaller features that will be added in a v1.x patch.

This is not for bugs (see #109 for that) and it is not for any particular implementation's feature completion status.

New Features

Add Two Permissible Values to the File Type element

Endpoint GET /file
Desired result File Type matches the File Type CDE in caDSR with 70 permissible values, including gVCF and HTSeq Count
Current version gVCF and HTSeq Count are not present as permissible values for file type
Is this a breaking change? Yes: response for File.type may now include new values which clients might not expect or be able to process.
Will implementations need to update? No: If implementation has no files with type gVCF or HTSeq Count this change does not affect them.
Date added June 24 2024
Current status feature not yet started as of June 24
Notes These permissible values were added to caDSR by request of the Federation (gVCF, HTSeq Count) subsequent to the File Type element being added to the API spec.

[Ssandor13]

New Features

Fix Missing Permissible Values and Remove 2 Permissible Values to the Disease Phase element

Endpoint GET /sample
Desired result caDSR Link: 12217251
Current version does not show Relapse/Refractory Remission Tumor Surveillance but shows Unknown and Not Reported
Is this a breaking change? Yes: response may now include new values which clients might not expect or be able to process.
Will implementations need to update? No: If implementation has no samples with types aforementioned above, this change does not affect them.
Date added July 8th, 2024
Current status feature not yet started as of July 8th, 2024

[GeoffLyle]

This is odd that the CDE would be changed after being released. @jknable Should we expect CDEs to change and continue to use the same version number?

[jknable]

The permissible values of CDE 12217251 (Disease Phase Timepoint Category) were last updated on 12/26/23 and 12/28/23 which removed Unknown and Not Reported and added 3 new permissible values (Relapsed/Refractory, Remission, Tumor Surveillance). Since the CDE had a status of “Draft New” at that time, no versioning was required.

CDE versioning rules require that once released, a CDE is versioned when a permissible value is removed, but not when new permissible value is added. There is a unique “Begin Date” attribute associated with each permissible value which can be used to identify updates.

[Ssandor13]

New Features

Remove 10 Permissible Values Tissue Preservation element

Endpoint GET /sample
Desired result caDSR Link: 5432521
Current version shows too many permissible values and does not match caDSR Link: 5432521
Will implementations need to update? Maybe: If implementation has no samples with types aforementioned in caDSR link, this change does not affect them, however, if the Federation used the wiki to map to then the implementation would need to be updated.
Date added July 8th, 2024
Current status feature not yet started as of July 8th, 2024

[GeoffLyle]

From the discussion on Preservation Method #64, we went with CDE 8028962 since it contains the EDTA permissible value.

[Ssandor13]

@GeoffLyle thank you. I apologize as I went with the link at the top of the discussion rather than reading through the entire discussion; I assumed that was the one we were using given it was at the top.

[asafieva]

CDE 6273393 v1.00 sample>>library_strategy
Two PVs were not included in wiki and in swagger.yaml "cde.v1.sample.LibraryStrategy": scMultiome, Spatial-tx.
Was this feature implemented as expected and Federation API needs only a subset of this CDE PVs? These PVs are not returned by any participating node implementations to the moment.
Date added July 18th, 2024
Current status feature not yet started as of July 18th, 2024

[Ssandor13]

@jknable Ethnicity was showing a lower case r instead of capital R for Not Reported permissible value. See snapshot below. There may be others with this issue as well. This will not only affect our wiki, mappings per Federation, but the API spec as well.

Date added: August 2nd, 2024
Current status: feature not yet started as of August 2nd, 2024.

[GeoffLyle]

New Features

Add 2 new CDEs accepted by the Data Harmonization team.

Library Source Material #114 - CDE 14808227
Library Selection Method #47 - CDE 6347743 v2

Note: For Library Selection Method we agreed to use a subset of the values in the CDE. These values are:

• Hybrid Selection
• Not applicable
• PCR
• Poly-A Enriched Genomic Library
• Random
• rRNA Depletion
• Unspecified

[GeoffLyle]

Note: Library Source Material #114 has been reopened for Data Harmonization team discussion on October 04, 2024.

[GeoffLyle]

Note: After discussion on October 18, 2024, Library Source Material #114 has been re-approved for implementation. Federation members will map blood tumor samples to "Bulk Tissue".

In the future the Federation would like to add a CDE that differentiates among "blood", "saliva", "tissue", "tears". (see #108 (comment))

[asafieva]

This is a question about sharing "study" information through CCDI API.
The only place to receive "study" information in API v1.0.0 is using the "namespace" endpoint. The component containing study attributes is "models.namespace.Metadata".
Besides KidsFirst, all other APIs nodes send "null" as namespace metadata, means no study information is sent.
Is there any plan to include "study" information into "namespace" endpoint API responses? Is there any other feasible way to connect "study" information to API entities such as "subject"?
CCDI hub and CPI extensively use study references.

Here is a consideration given by Brian Furner (PCDC) in the slack API channel.
It feels like information about studies could be a first order entity in a data model and right now they’re only represented as metadata about a namespace which is highly abstract.
In general, PCDC captures enrollment information for individual persons on one or more COG protocols, but what we have is just the study_id. It feels like we may want a discussion on how to handle reference data here as it is possible if not likely that there will be the same studies represented in different member source systems and unless we coordinate on this, search will be a challenge.

[GeoffLyle]

**New Features

Add 1 new CDEs accepted by the Data Harmonization team (October 4, 2024).**

Specimen Molecular Analyte Type #116 - CDE 15063661

Permissible values have been subset to:

• DNA
• Protein
• RNA

Note: Values 'Not Reported' and 'Unknown' are not currently included. This information should be known when sharing genomic data. If a member does not fill out this field, the API should return a null response. We believe not including these values will make this field easier for users to parse.