TCGAbiolinks download repeat clinical data and case count not the same with getProjectSummary

[wentgithub]

rm(list=ls()) library(TCGAbiolinks) library(dplyr) library(DT) query <- GDCquery(project = "TCGA-OV", data.category = "Clinical", file.type = "xml") GDCdownload(query) clinical <- GDCprepare_clinic(query, clinical.info = "patient") write.table(clinical, 'b.txt', sep="\t", row.names=F,col.names=T,quote=F)

you will see it has three duplicate bcr_patient_barcode

TCGA-3P-A9WA TCGA-59-A5PD TCGA-5X-AA5U ,waiting for your help, thanks a lot

============================================================================================================================ here is the resulr of TCGA-OV, you give clinical case_count is 608, but as the above getthing clinical data, even including the three repeat data, it is 590, not 608, so what is the real case count

[tiagochst]

I'll check the count and the duplicated sample. I did not touch that function for a long time.

[wentgithub]

thanks a lot, waiting for your reply

[tiagochst]

It seems TCGA-OV has 608 cases, but only 587 have clinical data. The numbers are the same from GDC data portal as shown below:

Example of case missing clinical data:

Code to check the samples missing: http://rpubs.com/tiagochst/TCGA-OV-cases

[wentgithub]

Thanks a lot. so what is the right way to get clinical data? query <- GDCquery(project = "TCGA-OV", data.category = "Clinical", file.type = "xml") GDCdownload(query) clinical <- GDCprepare_clinic(query, clinical.info = "patient") # and then dedup myself?

[tiagochst]

I suggest using the indexed data: clinical.indexed <- GDCquery_clinic(project = "TCGA-OV", type = "clinical")

[wentgithub]

it is really a pity, use the same code, I get not the same data as you

after write.table(clinical.indexed,"ov_clinical",sep="\t",row.names=F,col.names=T,quote=F) and you can see the result is double rownames, and most content is NA

[tiagochst]

Which version of TCGAbiolinks do you have installed? It seems it is an old one. Could you please update it from Github with:

withr::with_envvar(c(R_REMOTES_NO_ERRORS_FROM_WARNINGS="true"), 
  remotes::install_github('BioinformaticsFMRP/TCGAbiolinks')
)

[wentgithub]

I install from the bioconductor before. after installing from your suggestion, the counts now is right, but the content is of courese not I wanted at least, it lacks tumor stage information,

here is my code rm(list=ls()) library(TCGAbiolinks) library(dplyr) library(DT) clinical.indexed <- GDCquery_clinic(project = "TCGA-OV", type = "clinical")

where I am wrong, Thanks a lot

[tiagochst]

The indexed data is parsed from the XML files. It seems there is a problem with the parsing. You can get that information in the Biotab or XML.

query <- GDCquery(project = "TCGA-OV", 
                  data.category = "Clinical",
                  data.type = "Clinical Supplement", 
                  data.format = "BCR Biotab")
GDCdownload(query)
clinical.BCRtab.all <- GDCprepare(query)
clinical.BCRtab.all$clinical_patient_ov$tumor_grade

query <- GDCquery(project = "TCGA-OV", 
                  data.category = "Clinical",
                  data.type = "Clinical Supplement", 
                  data.format = "BCR Biotab")
GDCdownload(query)
clinical.BCRtab.all <- GDCprepare(query)
clinical.BCRtab.all$clinical_patient_ov$clinical_stage

[wentgithub]

I am so sorry, no matter install tcgabiolink from bioconductor or the method

withr::with_envvar(c(R_REMOTES_NO_ERRORS_FROM_WARNINGS="true"), remotes::install_github('BioinformaticsFMRP/TCGAbiolinks') ) you supplied.

running code

rm(list=ls()) library(TCGAbiolinks) library(dplyr) library(DT)

query <- GDCquery(project = "TCGA-OV", data.category = "Clinical", data.type = "Clinical Supplement", data.format = "BCR Biotab") GDCdownload(query) clinical.BCRtab.all <- GDCprepare(query)

both will report the same error so I check the function , here also says no argument sorry for disturbing you so many times, hope package TCGAbiolinks will become a more outstanding package for analysing tcga data