Why is TCGA Genomic Data Commons missing some data?

Question

[igor]

Last summer, TCGA data moved from TCGA Data Portal to the Genomic Data Commons. However, for some reason, some data is missing. At first, I thought it was a temporary glitch, but this hasn't changed over time.

For example, at GDAC Firebrowse, there are 194 TCGA-LAML 450K samples (2016-01-28 batch), but at GDC, there are 140 samples. There is also the GDC Legacy Archive for the legacy data, but that also contains 140 samples.

I've also seen similar discrepancies with RNA-seq data.

How can this be? This is a relatively old dataset. It should be stable. If there were any issue, I assume the samples would have been filtered years ago and it wouldn't be so many.

[GenoMax]

Have you tried emailing GDC support (support at nci-gdc.datacommons.io)?

[igor]

I haven't yet. Was hoping this would be faster and provide an unbiased answer.

[igor]

Apparently I was wrong about that. Got a nice response quickly:

There was a problem during the initial import of TCGA data that resulted in some of the TCGA-LAML being missed. Our development team is currently working on fixing this issue, but I'm not sure when it will be resolved.

[Zhenyu Zhang]

Btw, if you were in the loop during the old TCGA DCC era, there was an emergency redaction of TCGA LAML data for a couple of months. GDC happened to pull TCGA data from DCC and CGHub during that period of time, so that's why they are missing