using Firebrowser to identify disease type

Question

[dec986]

I'm trying to identify a disease cohort from TCGA, with the highest median co-expression of a certain drug. I'm using FireBrowser (http://firebrowse.org/) and the associated R package (https://rdrr.io/github/mariodeng/FirebrowseR/)

At first glance, the most appropriate command within the R package is Samples.mRNASeq but this command seems to do the opposite of what I need. Indeed, the website seems set up to work backwards from the way that I need, where the cohort is chosen first.

http://firebrowse.org/tutorial/FireBrowse-Tutorial.pdf doesn't seem to offer what I need.

None of the commands on https://rdrr.io/github/mariodeng/FirebrowseR/man/ are very helpful either.

TCGA seems to have ended in 2016, so I don't think that I'm going to get help from them.

thanks for your guidance

Answer 1

[dec986]

The solution to this is within the Samples.mRNASeq that gives data which can be saved in JSON format:

[0] {
        cohort                     "ACC",
        expression_log2            3.635731,
        gene                       "CD274",
        geneID                     29126,
        protocol                   "RSEM",
        sample_type                "TP",
        tcga_participant_barcode   "TCGA-PK-A5HB",
        z-score                    -0.01802174
    },
[1] {
        cohort                     "ACC",
        expression_log2            2.725785,
        gene                       "CD274",
        geneID                     29126,
        protocol                   "RSEM",
        sample_type                "TP",
        tcga_participant_barcode   "TCGA-PK-A5HA",
        z-score                    -0.1930097
    },...

the above data can be processed in a script to save to a hash/dict in Perl or whatever by cohort, and then you can use basic statistics to see which has the higher median expression