Questions about TCGA Pancreatic Adenocarcinoma Clinical Records

29 views
Skip to first unread message

EOIN O'HARE

unread,
May 12, 2022, 9:15:35 AM5/12/22
to cbiop...@googlegroups.com
Good morning,

I am a student at Rutgers University working on a project under the Rutgers Cancer Institute of New Jersey. For our project, we are using the clinical data of the combined study on Pancreatic Adenocarcinoma, which was combined from (TCGA, Firehose Legacy) and (TCGA, PanCancer Atlas). I had two questions about the clinical data. What is the naming convention of the patient id? Does it describe the institution or where the data was taken? For example, the 2J in TCGA-2J-AAB1-01. Also, is there a method of checking when each sample was taken? Thank you so much for your help!

Thank you so much,
Eoin O'Hare
Junior. B.S. Computer Engineering
Rutgers University Honors College

Tali Mazor

unread,
May 12, 2022, 9:29:54 AM5/12/22
to EOIN O'HARE, cbiop...@googlegroups.com
Hi Eoin,

The TCGA naming convention is described here: https://docs.gdc.cancer.gov/Encyclopedia/pages/TCGA_Barcode/
Within cBioPortal we also provide a clinical attirbute ('Tissue Source Site') which lists the site at which the sample was acquired.

In terms of when a sample was taken, it depends what you mean. As far as I know, the date of sample acquisition is not available. However, if your question relates to when a sample was taken relative to the patient disease course, that information is available for some patients, eg: https://www.cbioportal.org/patient?studyId=paad_tcga_pan_can_atlas_2018&caseId=TCGA-2J-AAB1  TCGA was primarily interested in analyzing pre-treatment tumors, so in general the samples are not post-treatment.

Finally, I advise against analyzing two different versions of the TCGA data together. This is because the Firehose Legacy and PanCancer Atlas studies contain many of the same patients/samples, which can bias your results. I suggest using just the PanCancer Atlas study, as it reflects a more recent re-analysis of all the data.

-Tali



--
You received this message because you are subscribed to the Google Groups "cBioPortal for Cancer Genomics Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cbioportal+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cbioportal/CAOUZJOAqjpzos675VEo9LMc_2yOcaJBiwoxtYTKtrQhG8yR2vA%40mail.gmail.com.

EOIN O'HARE

unread,
May 13, 2022, 3:25:08 AM5/13/22
to Tali Mazor, cbiop...@googlegroups.com
Good evening,

Thank you so much for the help! Hope you have a nice weekend!

Thanks again,

Eoin O'Hare
Junior. B.S. Computer Engineering
Rutgers University Honors College

Reply all
Reply to author
Forward
0 new messages