Matching transcriptomic data to clinical data

Question

[Khadija]

Hello everyone,

I'm new to the R language and I need some help.

I have downloaded the transcriptomic data related to the "TCGA-GBM" by using the package TCGAbiolinks, and I've also downloaded the clinical data (XML files), knowing that I'm interested in clinical_radiation.

What I want to do is for example combine the transcriptomic data to the clinical radiation for example.

Can anyone help me please? I'm really stuck.

Thanks in advance for your replies.

[Ram]

What do you mean by "combine"? Please show us the code you used to get your datasets.

[Khadija]

query_GBM2 <- GDCquery(project = "TCGA-GBM",
  data.category = "Transcriptome Profiling", 
  data.type = "Gene Expression Quantification",
  workflow.type="STAR - Counts")

GDCdownload(query_GBM2) data1<- GDCprepare(query_GBM2, save = TRUE, save.filename = "data1.rda")

GBMMatrice <- assay(data1,"tpm_unstrand")

Rnaseq_CorOutliers <- TCGAanalyze_Preprocessing(data1) 
dataNorm1 <- TCGAanalyze_Normalization(tabDF = data1, geneInfo = geneInfoHT) 
dataFilt1 <- TCGAanalyze_Filtering(tabDF = dataNorm1, method = "quantile", qnt.cut = 0.25) samplesNT1 <- TCGAquery_SampleTypes(barcode = colnames(dataFilt1), c("NT"))
samplesTP1 <- TCGAquery_SampleTypes(barcode = colnames(dataFilt1), c("TP")) 
dataDEGs1 <- TCGAanalyze_DEA(mat1 = dataFilt1[,samplesNT1], mat2 = dataFilt1[,samplesTP1], Cond1type = "Normal", Cond2type = "Tumor", fdr.cut = 0.01 , logFC.cut = 1, method = "glmLRT") dataDEGsFiltLevel1 <- TCGAanalyze_LevelTab(dataDEGs1,"Tumor","Normal", dataFilt1[,samplesTP1],dataFilt1[,samplesNT1]) 
query_GBM<- GDCquery( project = "TCGA-GBM", data.category = "Clinical", data.type = "Clinical Supplement", data.format = "BCR Biotab" ) 
GDCdownload(query_GBM) clinical.BCRtab.all <- GDCprepare(query_GBM) names(clinical.BCRtab.all) 
clinical.BCRtab.all$clinical_radiation_gbm %>% head %>% DT::datatable(options = list(scrollX = TRUE, keys = TRUE))

This is the code that I used. What I am trying to do is actually a transcriptomic analysis and retrieve the DEGs. After that, I have tried to combine the transcriptomic file (which contains the samples of patients and the genes) to the clinical data (especially the clinical radiation) and this is where I got stuck. As I said, I'm not familiar with R so I don't know if my procedure is right or not. Thank you for your reply.

[Ram]

Where did you get this code from? You'll probably just need a join of some sort to merge the clinical and transcriptomic data.

[Khadija]

I got it from the script of this topic: "TCGA Workflow: Analyze cancer genomics and epigenomics data using Bioconductor packages". It's the work of Tiago Chedraoui Silva and al. The fact is I don't know how I could merge the data. I think that what the two datasets have in common is the BCR_patient_barcode, but I don't know which function I could use...

[Khadija]

Here is the link to have access to the script:https://www.bioconductor.org/packages/release/workflows/vignettes/TCGAWorkflow/inst/doc/TCGAWorkflow.html

[Ram]

That helps, thanks. What is the output to:

head(GBMMatrice)
head(clinical.BCRtab.all$clinical_radiation_gbm)

[Khadija]

[Ram]

Please do not paste screenshots of plain text content, it is counterproductive. You can copy paste the content directly here (using the code formatting option shown below), or use a GitHub Gist if the content volume exceeds allowed length here.

[Khadija]

Ah, thank you, I didn't know. Here is the link for the GitHub Gist:

[Ram]

Nicely done. However, given that you have a ton of columns (which I did not know), maybe a better option would have been GBMMatrice[1:10,1:10]. In any case, please use gists/direct copy-paste in the future, it makes things a lot easier than screenshots

[Khadija]

Thanks for the advice! I'll keep that in mind for next time.

[Khadija]

[Khadija]

Thank you for your help. For the head (GBMMatrice), I have multiple samples not only the ones that are on the screenshot, but it's always with the same rows of genes.

[Ram]

Use a SummarizedExperiment object - that will work best for you. With some digging, you'll be able to figure out how to place both pieces of information in the same SummarizedExperiment object.

[Khadija]

Do you mean using this function like below? SummarizedExperiment(GBMMatrice) SummarizedExperiment(clinical.radiation)

[Ram]

No - please do some digging. The RNA data will be the assay and clinical metadata will become colData (I think, I'm not sure I got the keyword right). There is very minimal work required, I want you to put the effort in instead of asking around.

[Khadija]

Thank you for your help . I'll keep digging.

[Khadija]

I tried something else. I used the package TCGAutils to convert the file-id to the corresponding TCGAbarcode of the patients. Now, I'll try to use a function to merge the two datasets.

[Ram]

Try everything and post the solution that works for you as an answer. You are so close to cracking this!