Question: IDC API Availability

Hi Andrey,

I hope you are doing well.

I have a question about the IDC API. I am wondering if there is any specific planned date for its release?

Best regards,
Varik Hoang, BioDepot LLC.

@varikmp I understand the docs might have some inconsistencies, but indeed the IDC API has been released. I’ve just added the swagger link here: Endpoint Details - IDC User Guide.

Please take a look and share your experience. We very much are looking forward to any feedback about IDC API.

cc: @bill.clifford

Hi @fedorov,

Thank you so much for your response.

I followed up swagger link and tried out the endpoint /dicomMetadata. It worked great. Anyway, I am wondering if is there any extra parameter that I can filter rather than just the page_size. Let’s say crdc_study_uuid.

@varikmp do you mind if I move this conversation into the public section of the forum, so others can benefit from this discussion?

@fedorov Sure, no problem

1 Like

@bill.clifford can you please help @varikmp?

@varikmp, the /dicomMetadata endpoint is intended to support one specific user, so there are no additional parameters.
Perhaps if you tell us what you are trying to do we can help you.

@bill.clifford , thank you so much for your time getting back to me. For example, I have been trying to get all GS URL(s) of a specific patient with the ID. Let’s say 100_HMXXXXX. It is time-consuming to keep iterating over until the records found. The cohort file that I exported and saved from IDC did not have information of GS URL(s). I believe the author of this post Using API to get images for a particular TCGA patient barcode having the same issue that I have.

@varik, is this a manifest that was generated using one the API endpoints, /cohorts/manifest/{cohort_id} or /cohorts/manifest/preview? Both of those will optionally output the GCS URL of every instance in the cohort.

If you are exporting the manifest from the webapp, then note that the manifest output by the “Files” option does not include instance level metadata (SOPInstanceUIDs, GCS URLs). This is because there is a severe limit on the amount of data that can be downloaded.

However, the manifest output by the “BQ” option does include such instance level metadata. We generally recommend the BQ option. Such a manifest is also very useful because it can be joined with other IDC metadata tables for further investigations.


@varikmp I’ve just responded to that post. Does this address the issue you are having?