2.1 Useful information

2.2 Advanced search box

The following filters are available to select the data:

• Database
• Project filter
• Data category filter
• Data type filter
• Workflow filter
• Sample type filter
• Platform filter
• List of barcodes
• Clinical filter: gender, race, vital status, tumor stage.

2.3 Download & Prepare box

After the search step, the user can download the data and prepare it into an R object. The data can be prepared into one of two formats:

• SummarizedExperiment (Huber et al. 2015): An R object that contains the sample information matrix (clinical and subtype information), the assay matrix (i.e DNA methylation beta-values), the features matrix (i.e information for each probe in the DNA methylation assay matrix).
• Data Frame: A simple table with only the assay matrix (i.e Gene expression quantification)

In case a SummarizedExperiment object was selected the user will be able to add more metadata information into the sample information matrix. By checking “Add gistic2 and mutation information” and giving a list of genes, the GUI will download GISTIC2 data from the GDAC Firehose and Mutation Annotation Files (MAF) and their information to the object. This might be useful, for example, if the user wants to evaluate the survival of mutant samples vs non-mutant or if he wants to evaluate it on a heatmap, which can be used as labels.

Finally, the object will be saved with the same name as the file name field.

4.1 Useful information

4.2 Clinical data inconsistencies

# Get XML files and parse them
clin.query <- GDCquery(project = "TCGA-READ", 
                       data.category = "Clinical", 
                       file.type = "xml",
                       barcode = "TCGA-F5-6702")
GDCdownload(clin.query)
clinical.patient <- GDCprepare_clinic(clin.query, clinical.info = "patient")
clinical.patient.followup <- GDCprepare_clinic(clin.query, clinical.info = "follow_up")

# Get indexed data
clinical.index <- GDCquery_clinic("TCGA-READ")

select(clinical.patient,vital_status,days_to_death,days_to_last_followup) %>% datatable

select(clinical.patient.followup, vital_status,days_to_death,days_to_last_followup) %>% datatable

# Vital status should be the same in the follow up table 
filter(clinical.index,submitter_id == "TCGA-F5-6702") %>% select(vital_status,days_to_death,days_to_last_follow_up) %>% datatable

# Get XML files and parse them
recurrent.samples <- GDCquery(project = "TCGA-LIHC", 
                              data.category = "Transcriptome Profiling",
                              data.type = "Gene Expression Quantification", 
                              workflow.type = "HTSeq - Counts",
                              sample.type = "Recurrent Solid Tumor")$results[[1]] %>% select(cases)
recurrent.patients <- unique(substr(recurrent.samples$cases,1,12))
clin.query <- GDCquery(project = "TCGA-LIHC", 
                       data.category = "Clinical", 
                       file.type = "xml",
                       barcode = recurrent.patients)
GDCdownload(clin.query)
clinical.patient <- GDCprepare_clinic(clin.query, clinical.info = "patient") 

# Get indexed data
GDCquery_clinic("TCGA-LIHC") %>% filter(submitter_id %in% recurrent.patients) %>% 
    select(progression_or_recurrence,days_to_recurrence,tumor_grade) %>% datatable

# XML data
clinical.patient %>% select(bcr_patient_barcode,neoplasm_histologic_grade) %>% datatable

4.3 Tutorial Video

5.1 Mutation data search box

The available filters are:

• Tumor filter: a TCGA tumor, such as Breast invasive Carcinoma (TCGA-BRCA)
• Variant calling pipelines: Four separate variant calling pipelines are implemented for GDC data harmonization. Options: muse, varscan2, somaticsniper, mutect.

8.1 Sub-menu: DNA methylation

For DNA methylation raw data obtained in the form of Intensity Data (IDAT) files, we provide a pipeline using the R/Bioconductor minfi package (http://bioconductor.org/packages/minfi/) to process the data for downstream analysis (Aryee et al. 2014) performing a background and dye-bias correction with the preprocessnoob function followed by a detection P-value quality masking (sample specific) (Morris and Beck 2015) and probes overlapping repeats or single nucleotide polymorphisms masking (non-sample specific) (Zhou, Laird, and Shen 2016).

IDAT normalization menu: Table shows files identified that will be processed. Data retrieved from GEO (accession GSE61160).