Using combined hg19 and hg38 seed database

[Edward Yang]

Hi cBioPortal devs,

Thanks for the work put into cBioPortal!

Our raw data uses hg38 reference genome, but the public data (from datahub) uses hg19. I know that the recent recommendation is to use the combined hg19 and hg38 seed database to seed a new instance.

This is possibly a naive question: Does it make sense to have both of these types of studies in the one instance? Like, if I select a study that uses hg19 and another that uses hg38 and exploring them together through the "combined study" explorer, would the results make sense/be meaningful?

Thanks and I look forward to your response,

Ed Yang

Research Computing Engineer

WEHI

[Tim Kuijpers]

Dear Edward,

The GRCh37 (hg19) and GRCh38 (hg38) TCGA data versions are highly concordant, although there are some differences as explained in this paper by Gao et al. Hopefully, this can help you answer your question whether the results are meaningful. 

Please note that combining hg19 and hg38 studies is not fully supported yet, and will show a small warning when this combination is selected: 

' You are combining studies with molecular data based on both GRCh37/hg19 and GRCh38/hg38. This is not fully supported yet. Please double-check your findings.'

I hope this helps. If you have any further questions, please hit reply all, so our continued conversation is captured by the cBioPortal Google groups.

Best regards,

Tim

[Edward Yang]

Thanks for the clarification Tim!

Do you know when/if that feature will be added?

Thanks,

Ed Yang

Research Computing Engineer

WEHI