Skip to content

Clinical data

1. Cohort selection and view the data

A cohort can be selected in the top. Once selected, its details will be shown on the right.

Click "Reload your clinical datasets list" to refresh with any cohorts uploaded or edited since the page was loaded.

The data can be viewed as an interactive, sortable/searchable table in the "View the data" section, split across four tabs:

  • Gene expression: Since this table is usually large, by default it shows only the first 1000 rows to avoid overloading the server's memory. Switch to "Show everything" if you need the full table (may be slow for large datasets).
  • Survival: The patient survival/event data table (see Section 10.2.2 below for its format).
  • Meta data: The patient metadata table. You can also edit it directly here — toggle "Add a new metadata column" to create a new column and fill in its values, or "Delete a metadata column" to remove one.
  • Mutation data: The mutation data table, if one was uploaded for this cohort.

2. Survival analysis

This section examines the association between gene expression and survival outcomes within a selected patient cohort.

  1. Enter individual gene names line by line or select a custom gene set.
  2. Choose how to divide patients into high- and low-expression groups:
    • Using the median (default)
    • Using top 25% vs. bottom 25%
    • Using custom-defined thresholds (e.g., top X% vs. bottom Y%)
    • Manually specifying the two groups by entering sample names directly
  3. Choose sample filtering: "Use all samples" (default), or "Use the selected samples by a specific category" to restrict the analysis to patients matching one metadata category (e.g. a specific treatment group or subtype) before splitting them into groups.
  4. Select the event type
  5. Click "Start the survival analysis".
    • It generates a results table showing p-values and hazard ratios for each gene, sorted by the hazard ratio. The table is downloadable via "Download this table".
  6. Click on any row in the results table to display the corresponding Kaplan–Meier curve.
  7. Use the histogram feature to visualise gene expression distributions, which can help determine appropriate sample splitting criteria.

The survival events available for analysis (such as overall survival or progression-free survival) depend on the metadata included in your cohort dataset. For more details, please refer to 10.2. How to upload your own cohort section.

Adjustable graph parameters (Kaplan-Meier curve)
  • The size (width and height) of the figure.
  • The size of the X and Y axis/label font size.
  • The size of the legend title
  • The colour for the high- and low-expression group.
Adjustable graph parameters (Expression distribution histogram)
  • The size (width and height) of the figure.
  • X and Y axis/label font size, and graph title size.
  • The colour of the histogram bars.
  • Number of histogram bins (10-100, default: 20).
  • Option to switch from grey to white background.
Example Usage video

3. Gene correlation

This section computes and visualises pairwise correlations between the expression levels of multiple genes, across the patients in the selected cohort.

Enter two or more genes to get a correlation heatmap across all of them, then drill down into any specific pair to see its own scatter plot.

  1. Enter the genes to correlate, either by typing gene names line by line, or by toggling "Use the genes from the custom gene sets" to pull genes from one of your registered custom gene sets instead.
  2. Choose sample filtering: "Use all samples" (default), or "Use the selected samples by a specific category" to restrict the analysis to patients matching one metadata category.
  3. Choose the calculation method: Pearson or Spearman.
  4. Click "Calculate the correlation" to compute the full pairwise correlation matrix across all your input genes.
  5. View the results:
    • Correlation table: Select one of your input genes from the "Select a gene to show its correlation result table" drop-down to see a table of that gene's correlation (r-value and p-value) against every other gene you entered. The table is downloadable via "Download this table".
    • Pairwise heatmap: By default, the "Plot" panel shows a clustered heatmap of the full pairwise correlation matrix across all your input genes — genes are automatically reordered by hierarchical clustering so similarly-correlated genes sit together.
    • Scatter plot for one pair: Once you've selected a gene from the drop-down above, click its row in the correlation table to pick a second gene — the "Plot" panel then switches to a scatter plot of just that gene pair, with an optional regression line.
Adjustable graph parameters (Pairwise heatmap)
  • Figure width and height
  • X/Y axis label size (set to 0 to hide the gene name labels entirely)
  • Legend font size
  • Colour for the lowest correlation (-1), highest correlation (+1), and mid correlation (0)
Adjustable graph parameters (Scatter plot)
  • Figure width and height
  • X/Y axis label size and title size
  • Colour of the dots
  • Option to show the correlation line (on by default)
  • Option to use a white background
Example Usage video

4. Mutation analysis

This section helps you analyse gene mutations in the selected cohort.

Mutant and wild type (WT). For each gene, a sample is classified as "Mutant" when the mutation data contains any record of that gene for the sample, and as "WT" otherwise. A sample without any record in the mutation data is therefore WT. The samples used differ between the tabs:

Tab Samples used Mutant WT
Frequency Plot Samples in the mutation data Samples with any record of the gene —
Survival analysis Samples in both the gene expression data and the survival data Samples with any record of the gene The other samples
Gene expression plot Samples in the gene expression data Samples with any record of the gene The other samples
  • When the samples are filtered by a metadata category (step 2 below), only the samples in that category are used in all the tabs.
  • Mutation frequency (%) = the number of samples with any record of the gene ÷ the number of samples in the mutation data × 100 (after the sample filtering). The gene expression and survival data are not used here. As a sample without any mutation does not appear in the mutation data, the denominator is the number of samples with at least one reported mutation (for the pre-installed TCGA cohorts, practically all the samples analysed by whole-exome sequencing).
  • In the Survival analysis and Gene expression plot tabs, samples in the gene expression data that were not analysed for mutations are also treated as WT. When uploading your own cohort, please make sure that all the samples in the gene expression data were analysed for mutations.

  • For the pre-installed TCGA cohorts, the gene expression data contains one primary tumour sample per patient (see 10.1), so normal tissue samples are not included in the WT group.

  • Input the genes to investigate by:

    • Entering gene names one per line in the text box ('Text input')
    • Choosing to analyse all genes in the dataset ('Use all genes')
    • Selecting a gene set from a custom gene set ('Select from custom genesets')
  • Filter the samples if necessary
    • By default, all cohort samples are included ('Use all samples')
    • When choosing 'Use the selected samples by a specific category', you can filter samples using metadata (treatment group, cancer subtype, demographics) for targeted analysis
  • Click "Calculate the mutation frequency". A Results table will appear in the bottom left section. The following plots will be generated in the tabs:
    • Frequency Plot tab: A bar plot showing the frequency or the counts of mutated genes. The Y-axis displays either mutation count or frequency (%). You can adjust the number of genes to display.
    • Survival analysis tab: Select an event for survival analysis. Clicking a gene name displays a Kaplan-Meier curve comparing the wild-type patients and the mutant patients.
    • Gene expression plot tab: This comparison examines the expression levels between the wild-type and mutant groups. Choose a plot type — Box plot, Violin plot, Swarm plot, or Violin + Swarm plot.
      1. Click a gene in the mutation analysis results table
      2. Enter genes in the input field to generate a table listing these genes
      3. Click a gene from the Input genes table to generate a plot comparing its expression between the wild-type and mutant patients
Adjustable graph parameters
  1. Frequency Plot
    • Plot size (width and height)
    • X and Y label size
    • Legend size
    • Score font size (the value labels drawn on each bar)
    • Colour for the highest value, and for zero
    • Number of genes to display
    • White background option
    • Option to hide scores on each bar
  2. Survival analysis
    • Figure size (width and height)
    • X and Y axis/label font size
    • Legend title size
    • Colours for high and low expression groups
  3. Gene expression plot
    • Plot size (width and height)
    • X and Y label size
    • Dot size (for the swarm plot)
    • Graph title and legend font size
    • Colours for wild-type and mutant groups
    • White background option
Example Usage video

5. Gene expression across subtypes

When metadata for the cohort is provided and patients can be divided into subtypes, users can compare gene expression across patient subgroups.

  1. Enter the genes or select a custom gene set
  2. Select a category for subtype from the "Group by" drop-down menu. If the category has more than two subtypes, you can toggle "Want to compare only two subtypes?" to restrict the comparison to exactly two of them.
  3. Click "Start comparing" to compare gene expression across subtypes. Note that visualisation may be slow and cluttered when there are many subtypes in the selected group.
  4. A result table with statistical scores and p-values will be generated (downloadable via "Download this table"). Statistical scores include W values for two subtypes and H values for three or more subtypes.
  5. Clicking any row in the table displays a visualisation on the right. Available plot types include Box plot, Violin plot, Swarm plot, or Violin + Swarm plot
Adjustable graph parameters
  • Figure size (width and height)
  • X and Y axis/label font size
  • Graph title size
  • Dot size (for the swarm plot)
  • Colour palette for each subtype, or "Use a single colour" to override it with one flat colour
  • Option to rotate the X-axis labels
  • Option to switch from grey to white background
Example Usage video

6. Signature analysis

This section performs signature analysis on gene expression data from the selected cohort to evaluate the activity or presence of specific biological processes.

  1. Choose the input type by either:
    • Selecting from custom gene sets
    • Entering gene names line by line
  2. Choose sample filtering: "Use all samples" (default), or "Use the selected samples by a specific category" to restrict the analysis to patients matching one metadata category.

  3. Select the calculation method:

    • GSVA (Gene Set Variation Analysis) or ssGSEA (single-sample Gene Set Enrichment Analysis) is available.

  4. Click "Calculate the signature score". This generates a result table with scores for each sample, downloadable via "Download this table".

  5. Three plots are generated:
    • Survival analysis plot tab:
      • Generates a Kaplan-Meier plot.
      • Allows selection of methods to split samples into high and low-score patients: by median, Top 25% vs Bottom 25%, or a custom Top X% vs Bottom Y%.
    • Score comparison tab:
      • Select the group to compare signature scores. If the group has more than two subtypes, you can toggle "Want to compare only two subtypes?" to restrict the comparison to exactly two of them.
      • Click "Show a plot". Four plot types are available (Box plot, Violin plot, Swarm plot, and Violin + Swarm plot).
    • Distribution tab:
      • Generates a histogram of signature scores to help determine appropriate sample splitting criteria.
What is GSVA and ssGSEA?

GSVA (Gene Set Variation Analysis) calculates an enrichment score for each gene set by transforming gene expression data into a pathway activity score across samples. It uses kernel-based density estimation to assess the relative enrichment of a gene set, comparing it to the overall expression distribution in the dataset.

ssGSEA (single-sample Gene Set Enrichment Analysis), on the other hand, ranks genes within each sample and calculates an enrichment score based on the ranked positions of genes in a gene set. It evaluates how consistently genes of a set are positioned at the top or bottom of the ranked gene list for each individual sample.

Adjustable graph parameters
  • Survival analysis plot
    • Figure size (width and height)
    • X and Y axis/label font size
    • Legend title size
    • Colors for high and low expression groups
  • Score comparison plot
    • Figure size (width and height)
    • X and Y axis/label font size and graph title font size
    • Dot size (for the swarm plot)
    • Color palette, or "Use a single colour" to override it with one flat colour
    • Option to rotate the X-axis labels
    • Option to switch from grey to white background
  • Histogram
    • Figure size (width and height)
    • X and Y axis/label font size
    • Histogram bin color
    • Number of histogram bins
    • Option to switch from grey to white background
Example Usage video

7. Deconvolution analysis

This section provides deconvolution analysis from patients' gene expression data (typically bulk RNAseq). While several deconvolution tools exist, two are available here: MCPcounter and xCell.

  1. Select the cohort
  2. Choose either MCPcounter or xCell as your method, then click "Start deconvolution". A deconvolution result table will appear on the right, downloadable via "Download this table".

The interface provides three additional analysis tabs.

Example Usage video

7.1. Generating a heatmap/barplot

First, you can generate a heatmap or barplot to visualise cell type fractions across samples.

  1. Choose which samples to use:
    • "All samples": Includes every sample in the dataset
    • "Filter from metadata":
      • Filter samples by subtypes using information from the metadata files
      • The number of samples selected will be displayed after choosing a category
    • "Text input": Enter sample IDs (patient IDs) line by line in the text box, ensuring no extra spaces
  2. Choose which cell types to include:
    • "All cell types": Includes all available cell types
    • "Select cell types":
      • A table of available cell types will be displayed
      • Click on specific cell types you want to include in the plot
  3. Click "Show a heatmap" to generate a heatmap and barplot on the right
Adjustable graph parameters (Heatmap)
  • Figure size (width and height)
  • X and Y axis font size
  • Legend font size
  • Colour for the highest value, and for zero
Adjustable graph parameters (Barplot)
  • Figure size (width and height)
  • X and Y axis font size
  • Legend font size
  • "Percentile plot" option: switch from raw scores to percentile-scaled values

7.2. Exploring correlations between gene expression and cell type abundance

This feature allows you to analyse the relationship between specific gene expression levels and cell type abundance.

  1. Enter the gene names line by line (or choose a gene set)
  2. Select a cell type to investigate
  3. Choose sample filtering: "Use all samples" (default), or "Use the selected samples by a specific category" to restrict the analysis to patients matching one metadata category.
  4. Choose a correlation method, either Pearson or Spearman
  5. Click "Calculate the correlation". This calculates the correlation between gene expression and cell type abundance, generating a table with correlation coefficients and p-values, downloadable via "Download this table".
  6. Clicking any row in the result table generates a scatter plot.
Adjustable graph parameters
  • Figure size (width and height)
  • X and Y axis font size
  • Legend font size
  • Dot and correlation line colours
  • Option to display or hide the correlation line
  • Option to use a white background

7.3. Association between mutations and cell type abundance

This feature allows you to compare the abundance of a specific cell type between samples with and without mutations (wild type, WT) in each gene. All samples in the deconvolution result are used. A sample is classified as "Mutant" for a gene when the mutation data contains any record of that gene for the sample; otherwise, it is classified as "WT" (as in Mutation analysis).

  1. Enter the gene names line by line (or choose a gene set)
  2. Select a cell type to investigate
  3. Click "Start comparing". For each gene, the abundance of the selected cell type is compared between the Mutant and WT groups using the Wilcoxon rank-sum test. A result table (gene, number of mutant/WT samples, statistic, and p-value) sorted by p-value is generated, downloadable via "Download this table".
    • Genes without mutated samples are listed above the table.
    • If none of the input genes have mutated samples, no table is generated.
    • If the cohort has no mutation data, a message is shown.
  4. Clicking any row in the result table generates a box plot (default), violin plot, swarm plot, or violin + swarm plot.
  5. Turn on "Split the mutant group by the type of mutation" to show only the mutant samples, split by amino acid change (the Amino_Acid_Change column in TCGA data), as a swarm plot. Samples with several amino acid changes in the gene are grouped as "Multiple". Only the top X most frequent amino acid changes (10 by default, adjustable) are shown, and the rest are grouped as "Others". If the mutation data does not contain the type of mutation for the selected gene, an error message is shown and the switch is turned off.
Adjustable graph parameters
  • Figure size (width and height)
  • Dot size (swarm plot)
  • X and Y axis label/title font size, and graph title font size
  • Colours for the mutant and wild type groups
  • Colour palette for the mutation types
  • Option to use a white background
  • Option to rotate the X axis labels

8. Compare cohorts

In this section, you can compare gene expression or mutation frequency across different cohorts.

  1. (You do not have to select a cohort in this section)
  2. Enter gene names line by line, or choose a custom gene set
  3. The list of genes will appear. Click the gene you want to investigate.
  4. Select the cohorts you want to include.
    • All cohorts stored in OmicsBridge will be listed.
    • You can select multiple cohorts, but note that generating figures may take longer, especially for gene expression analysis. (ex. When selecting all TCGA cohorts, it takes ~30 sec for the mutation frequency analysis and ~2 minutes for the gene expression analysis.)
  5. Click "Compare mutation frequencies" or "Compare gene expressions", depending on the tab you're in ("Mutation Frequency" or "Gene expression")
    • In "Mutation Frequency," a bar plot will display the number or percentage of patients with mutations in the selected gene across the chosen cohorts — choose which via the "Y axis" option.
      • The percentage is calculated for each cohort in the same way as in 4. Mutation analysis: the number of samples with any record of the gene in the cohort's mutation data ÷ the number of samples in that mutation data × 100.
      • Cohorts without mutation data, and cohorts without any record of the gene in their mutation data, are not shown in the plot.
    • In "Gene expression," a box plot is generated that compares gene expression levels across the selected cohorts
Adjustable graph parameters (Mutation Frequency)
  • Figure size (width and height)
  • Font size for x-axis, y-axis, legend, and the score labels on each bar
  • Colour for the highest value, and for zero
  • Option to use a white background
  • Option to hide the score labels on each bar
Adjustable graph parameters (Gene expression)
  • Figure size (width and height)
  • Font size for x-axis and y-axis
  • Option to use a white background
Example Usage video

9. Cancer Gene Census (COSMIC)

OmicsBridge includes a database of cancer predisposition genes sourced from Cancer Gene Census from COSMIC. This feature helps you identify which genes from your input are known to be associated with cancer predisposition.

  1. Enter gene names line by line, or toggle "Use the genes from the custom gene sets" to pull genes from one of your registered custom gene sets instead.
  2. If any of the genes you entered are associated with cancer predisposition, they will appear in the results table. If none match, the complete database will be displayed instead. The table is downloadable via "Download this table".
Example Usage video

10. Manage the cohort database

The users can manage the cohort database and upload or delete datasets in the “Cohort database” tab.

10.1. Pre-installed cohort

TCGA data (34 cancer types, see the table below) is available as a separate, optional download — a package you download from Zenodo and extract into the OmicsBridge directory yourself, rather than data bundled with the software installation. It includes mRNA sequencing results, clinical information, metadata and mutation data downloaded from UCSC Xena, with gene expression values transformed as log2(RSEM normalised count+1). The gene expression data was curated to contain one tumour sample per patient (normal tissue samples are not included) and only genes with a gene symbol (see "Data source and version details" below).

Data source and version details

Gene expression, clinical/metadata, and survival data were downloaded from the TCGA Hub; mutation data was downloaded separately from the GDC Hub — two different data hubs on UCSC Xena.

File Xena hub Dataset Version
Gene expression TCGA Hub gene expression RNAseq > Illumina Hiseq 2017-10-13 (2017-09-08 for the 3 combined cohorts)
Clinical / metadata TCGA Hub phenotype > Curated survival data 2019-12-06
Survival TCGA Hub phenotype > Phenotypes 2018-09-13 (2020-05-15 for TCGA_LUNG)
Mutation GDC Hub somatic mutation > mutect2 snv (deprecated) 2019-07-19

The mutation files originally used a 16-character sample ID format (with a trailing sample-portion letter, e.g. TCGA-2F-A9KO-01A), while the other three files use a 15-character format (e.g. TCGA-2F-A9KO-01). This trailing character was removed from the mutation files’ sample IDs so that samples can be matched across all four data types; no other value was changed.

This particular MuTect2 mutation dataset has since been marked deprecated on the GDC Xena Hub and replaced there by a newer "Ensemble Somatic Variant" pipeline output. The mutation data included here reflects the original MuTect2 release as downloaded in 2019, not the current GDC release.

The gene expression files (HiSeqV2) originally contained several sample types per cohort — primary tumour, solid tissue normal, recurrent tumour, metastatic and additional samples — and some patients had more than one sample. Since the survival data also has a row for each of these samples, the same patient would be counted more than once in the analyses (e.g. once as a tumour and once as a normal tissue sample). The gene expression files were therefore curated as follows (October 2026, 11,966 → 10,999 samples in total); no value was changed:

  • One tumour sample per patient. Only primary tumour samples (sample type code 01, or 03 "Primary Blood Derived Cancer" for TCGA_LAML) were kept. For TCGA_SKCM, metastatic samples (06) were also kept because most SKCM samples are metastatic. When a patient (the first 12 characters of the sample ID, e.g. TCGA-2F-A9KO) still had more than one sample, the primary tumour sample was kept.
  • Removed samples. Solid tissue normal (11), recurrent tumour (02), additional new primary (05), additional metastatic (07) and metastatic samples (06) outside TCGA_SKCM. Patients without a primary tumour sample in the HiSeqV2 dataset are therefore not included (e.g. 15 patients in TCGA_COAD with only a normal tissue sample).
  • Removed genes. The 29 genes without a gene symbol (e.g. ?|26823), leaving 20,501 genes.

As normal tissue samples are not included, tumour vs. normal tissue comparisons are not possible with these cohorts. The survival, metadata and mutation files were not changed; samples are matched to the gene expression data by sample ID.

Number of samples per cohort (before → after curation)
Cohort Before After Cohort Before After
TCGA_ACC 79 79 TCGA_LIHC 423 371
TCGA_BLCA 426 407 TCGA_LUAD 576 515
TCGA_BRCA 1218 1097 TCGA_LUNG 1129 1017
TCGA_CESC 308 303 TCGA_LUSC 553 502
TCGA_CHOL 45 36 TCGA_MESO 87 87
TCGA_COAD 329 286 TCGA_PAAD 183 178
TCGA_COADREAD 434 380 TCGA_PCPG 187 179
TCGA_DLBC 48 48 TCGA_PRAD 550 497
TCGA_ESCA 196 184 TCGA_READ 105 94
TCGA_GBM 172 154 TCGA_SARC 265 259
TCGA_GBMLGG 702 670 TCGA_SKCM 474 470
TCGA_HNSC 566 520 TCGA_TGCT 156 150
TCGA_KICH 91 66 TCGA_THCA 572 505
TCGA_KIRC 606 533 TCGA_THYM 122 120
TCGA_KIRP 323 290 TCGA_UCEC 201 176
TCGA_LAML 173 173 TCGA_UCS 57 57
TCGA_LGG 530 516 TCGA_UVM 80 80
Download commands (for full reproducibility)

These are the exact commands used to assemble this dataset collection, provided so it can be independently verified, updated, or reproduced.

Gene expression — TCGA Hub, HiSeqV2 dataset:

wget -O TCGA_ACC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.ACC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_BLCA/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.BLCA.sampleMap%2FHiSeqV2.gz
wget -O TCGA_BRCA/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.BRCA.sampleMap%2FHiSeqV2.gz
wget -O TCGA_CESC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.CESC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_CHOL/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.CHOL.sampleMap%2FHiSeqV2.gz
wget -O TCGA_COAD/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.COAD.sampleMap%2FHiSeqV2.gz
wget -O TCGA_DLBC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.DLBC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_ESCA/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.ESCA.sampleMap%2FHiSeqV2.gz
wget -O TCGA_GBM/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.GBM.sampleMap%2FHiSeqV2.gz
wget -O TCGA_HNSC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.HNSC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_KICH/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.KICH.sampleMap%2FHiSeqV2.gz
wget -O TCGA_KIRC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.KIRC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_KIRP/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.KIRP.sampleMap%2FHiSeqV2.gz
wget -O TCGA_LAML/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LAML.sampleMap%2FHiSeqV2.gz
wget -O TCGA_LGG/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LGG.sampleMap%2FHiSeqV2.gz
wget -O TCGA_LIHC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LIHC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_LUAD/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LUAD.sampleMap%2FHiSeqV2.gz
wget -O TCGA_LUSC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LUSC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_MESO/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.MESO.sampleMap%2FHiSeqV2.gz
wget -O TCGA_PAAD/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.PAAD.sampleMap%2FHiSeqV2.gz
wget -O TCGA_PCPG/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.PCPG.sampleMap%2FHiSeqV2.gz
wget -O TCGA_PRAD/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.PRAD.sampleMap%2FHiSeqV2.gz
wget -O TCGA_READ/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.READ.sampleMap%2FHiSeqV2.gz
wget -O TCGA_SARC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.SARC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_SKCM/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.SKCM.sampleMap%2FHiSeqV2.gz
wget -O TCGA_TGCT/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.TGCT.sampleMap%2FHiSeqV2.gz
wget -O TCGA_THCA/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.THCA.sampleMap%2FHiSeqV2.gz
wget -O TCGA_THYM/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.THYM.sampleMap%2FHiSeqV2.gz
wget -O TCGA_UCEC/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.UCEC.sampleMap%2FHiSeqV2.gz
wget -O TCGA_UCS/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.UCS.sampleMap%2FHiSeqV2.gz
wget -O TCGA_UVM/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.UVM.sampleMap%2FHiSeqV2.gz
wget -O TCGA_COADREAD/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.COADREAD.sampleMap%2FHiSeqV2.gz
wget -O TCGA_GBMLGG/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.GBMLGG.sampleMap%2FHiSeqV2.gz
wget -O TCGA_LUNG/Gene_expression_RSEM.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LUNG.sampleMap%2FHiSeqV2.gz

Clinical / metadata — TCGA Hub, clinicalMatrix dataset:

wget -O TCGA_ACC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.ACC.sampleMap%2FACC_clinicalMatrix
wget -O TCGA_BLCA/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.BLCA.sampleMap%2FBLCA_clinicalMatrix
wget -O TCGA_BRCA/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.BRCA.sampleMap%2FBRCA_clinicalMatrix
wget -O TCGA_CESC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.CESC.sampleMap%2FCESC_clinicalMatrix
wget -O TCGA_CHOL/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.CHOL.sampleMap%2FCHOL_clinicalMatrix
wget -O TCGA_COAD/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.COAD.sampleMap%2FCOAD_clinicalMatrix
wget -O TCGA_DLBC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.DLBC.sampleMap%2FDLBC_clinicalMatrix
wget -O TCGA_ESCA/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.ESCA.sampleMap%2FESCA_clinicalMatrix
wget -O TCGA_GBM/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.GBM.sampleMap%2FGBM_clinicalMatrix
wget -O TCGA_HNSC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.HNSC.sampleMap%2FHNSC_clinicalMatrix
wget -O TCGA_KICH/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.KICH.sampleMap%2FKICH_clinicalMatrix
wget -O TCGA_KIRC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.KIRC.sampleMap%2FKIRC_clinicalMatrix
wget -O TCGA_KIRP/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.KIRP.sampleMap%2FKIRP_clinicalMatrix
wget -O TCGA_LAML/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LAML.sampleMap%2FLAML_clinicalMatrix
wget -O TCGA_LGG/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LGG.sampleMap%2FLGG_clinicalMatrix
wget -O TCGA_LIHC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LIHC.sampleMap%2FLIHC_clinicalMatrix
wget -O TCGA_LUAD/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LUAD.sampleMap%2FLUAD_clinicalMatrix
wget -O TCGA_LUSC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LUSC.sampleMap%2FLUSC_clinicalMatrix
wget -O TCGA_MESO/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.MESO.sampleMap%2FMESO_clinicalMatrix
wget -O TCGA_PAAD/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.PAAD.sampleMap%2FPAAD_clinicalMatrix
wget -O TCGA_PCPG/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.PCPG.sampleMap%2FPCPG_clinicalMatrix
wget -O TCGA_PRAD/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.PRAD.sampleMap%2FPRAD_clinicalMatrix
wget -O TCGA_READ/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.READ.sampleMap%2FREAD_clinicalMatrix
wget -O TCGA_SARC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.SARC.sampleMap%2FSARC_clinicalMatrix
wget -O TCGA_SKCM/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.SKCM.sampleMap%2FSKCM_clinicalMatrix
wget -O TCGA_TGCT/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.TGCT.sampleMap%2FTGCT_clinicalMatrix
wget -O TCGA_THCA/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.THCA.sampleMap%2FTHCA_clinicalMatrix
wget -O TCGA_THYM/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.THYM.sampleMap%2FTHYM_clinicalMatrix
wget -O TCGA_UCEC/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.UCEC.sampleMap%2FUCEC_clinicalMatrix
wget -O TCGA_UCS/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.UCS.sampleMap%2FUCS_clinicalMatrix
wget -O TCGA_UVM/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.UVM.sampleMap%2FUVM_clinicalMatrix
wget -O TCGA_COADREAD/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.COADREAD.sampleMap%2FCOADREAD_clinicalMatrix
wget -O TCGA_GBMLGG/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.GBMLGG.sampleMap%2FGBMLGG_clinicalMatrix
wget -O TCGA_LUNG/meta.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/TCGA.LUNG.sampleMap%2FLUNG_clinicalMatrix

Survival — TCGA Hub, survival dataset:

wget -O TCGA_ACC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FACC_survival.txt
wget -O TCGA_BLCA/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FBLCA_survival.txt
wget -O TCGA_BRCA/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FBRCA_survival.txt
wget -O TCGA_CESC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FCESC_survival.txt
wget -O TCGA_CHOL/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FCHOL_survival.txt
wget -O TCGA_COAD/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FCOAD_survival.txt
wget -O TCGA_DLBC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FDLBC_survival.txt
wget -O TCGA_ESCA/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FESCA_survival.txt
wget -O TCGA_GBM/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FGBM_survival.txt
wget -O TCGA_HNSC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FHNSC_survival.txt
wget -O TCGA_KICH/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FKICH_survival.txt
wget -O TCGA_KIRC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FKIRC_survival.txt
wget -O TCGA_KIRP/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FKIRP_survival.txt
wget -O TCGA_LAML/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FLAML_survival.txt
wget -O TCGA_LGG/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FLGG_survival.txt
wget -O TCGA_LIHC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FLIHC_survival.txt
wget -O TCGA_LUAD/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FLUAD_survival.txt
wget -O TCGA_LUSC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FLUSC_survival.txt
wget -O TCGA_MESO/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FMESO_survival.txt
wget -O TCGA_PAAD/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FPAAD_survival.txt
wget -O TCGA_PCPG/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FPCPG_survival.txt
wget -O TCGA_PRAD/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FPRAD_survival.txt
wget -O TCGA_READ/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FREAD_survival.txt
wget -O TCGA_SARC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FSARC_survival.txt
wget -O TCGA_SKCM/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FSKCM_survival.txt
wget -O TCGA_TGCT/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FTGCT_survival.txt
wget -O TCGA_THCA/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FTHCA_survival.txt
wget -O TCGA_THYM/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FTHYM_survival.txt
wget -O TCGA_UCEC/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FUCEC_survival.txt
wget -O TCGA_UCS/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FUCS_survival.txt
wget -O TCGA_UVM/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FUVM_survival.txt
wget -O TCGA_COADREAD/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FCOADREAD_survival.txt
wget -O TCGA_GBMLGG/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FGBMLGG_survival.txt
wget -O TCGA_LUNG/survival.tsv https://tcga-xena-hub.s3.us-east-1.amazonaws.com/download/survival%2FLUNG_survival.txt

Mutation — GDC Hub, mutect2_snv dataset (MuTect2 calls; not available for the 3 combined cohorts):

wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-ACC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-BLCA.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-BRCA.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-CESC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-CHOL.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-COAD.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-DLBC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-ESCA.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-GBM.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-HNSC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-KICH.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-KIRC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-KIRP.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-LAML.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-LGG.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-LIHC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-LUAD.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-LUSC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-MESO.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-PAAD.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-PCPG.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-PRAD.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-READ.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-SARC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-SKCM.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-TGCT.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-THCA.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-THYM.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-UCEC.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-UCS.mutect2_snv.tsv.gz
wget https://gdc-hub.s3.us-east-1.amazonaws.com/download/TCGA-UVM.mutect2_snv.tsv.gz

Sample ID reformatting applied to mutations.tsv in 29 of the 31 cohorts (TCGA_ACC and TCGA_LUAD already used the matching 15-character format and were left unchanged) — the last character of every value in the sample column was removed:

awk -F'\t' -v OFS='\t' 'NR==1{print; next} {$1=substr($1,1,length($1)-1); print}' mutations.tsv > mutations_tmp.tsv && mv mutations_tmp.tsv mutations.tsv

Gene expression curation applied to Gene_expression_RSEM.tsv in all 34 cohorts (R, with the data.table package; values are read and written as text, so they are not changed):

library(data.table)
ex <- fread('Gene_expression_RSEM.tsv', colClasses = 'character', check.names = FALSE)
kept_types <- c('01', '03')                              # c('01', '03', '06') for TCGA_SKCM
samples <- colnames(ex)[-1]
candidates <- samples[substr(samples, 14, 15) %in% kept_types]
candidates <- candidates[order(match(substr(candidates, 14, 15), kept_types))]
chosen <- candidates[!duplicated(substr(candidates, 1, 12))]  # one sample per patient
kept <- samples[samples %in% chosen]                         # keep the original column order
ex <- ex[!startsWith(ex[[1]], '?|'), c(colnames(ex)[1], kept), with = FALSE]
fwrite(ex, 'Gene_expression_RSEM.tsv', sep = '\t', quote = FALSE)
TCGA abbreviation
Abbreviation Cancer type
TCGA_ACC Adrenocortical carcinoma
TCGA_BLCA Bladder Urothelial Carcinoma
TCGA_BRCA Breast invasive carcinoma
TCGA_CESC Cervical squamous cell carcinoma and endocervical adenocarcinoma
TCGA_CHOL Cholangiocarcinoma
TCGA_COAD Colon adenocarcinoma
TCGA_DLBC Lymphoid Neoplasm Diffuse Large B-cell Lymphoma
TCGA_ESCA Esophageal carcinoma
TCGA_GBM Glioblastoma multiforme
TCGA_HNSC Head and Neck squamous cell carcinoma
TCGA_KICH Kidney Chromophobe
TCGA_KIRC Kidney renal clear cell carcinoma
TCGA_KIRP Kidney renal papillary cell carcinoma
TCGA_LAML Acute Myeloid Leukemia
TCGA_LGG Brain Lower Grade Glioma
TCGA_LIHC Liver hepatocellular carcinoma
TCGA_LUAD Lung adenocarcinoma
TCGA_LUSC Lung squamous cell carcinoma
TCGA_MESO Mesothelioma
TCGA_PAAD Pancreatic adenocarcinoma
TCGA_PCPG Pheochromocytoma and Paraganglioma
TCGA_PRAD Prostate adenocarcinoma
TCGA_READ Rectum adenocarcinoma
TCGA_SARC Sarcoma
TCGA_SKCM Skin Cutaneous Melanoma
TCGA_TGCT Testicular Germ Cell Tumors
TCGA_THCA Thyroid carcinoma
TCGA_THYM Thymoma
TCGA_UCEC Uterine Corpus Endometrial Carcinoma
TCGA_UCS Uterine Carcinosarcoma
TCGA_UVM Uveal Melanoma
TCGA_COADREAD Colon and Rectal Cancer
TCGA_GBMLGG lower grade glioma and glioblastoma
TCGA_LUNG Lung Cancer

10.2. How to upload your own cohort

The users can upload their own cohort and analyse it here. Three files (Gene expression, Patient survival information, and Metadata) should be uploaded. Optionally, mutation data can be added. Each data has to follow the following data format.

10.2.1 Gene expression

A tab-delimited table of the gene expression of each sample (genes × samples(patients)) from bulk RNAseq (or microarray).

  • We recommend uploading normalised, log2-transformed values, e.g. log2(TPM+1) or log2(FPKM+1). OmicsBridge does not normalise the data for you; the values are used exactly as uploaded.
  • Raw counts can also be uploaded and every analysis will still run, but the results can be misleading (see "About normalisation" below).
  • Rows (index): gene names.
  • Columns (headers): sample names that match those used in your clinical data and metadata.
About normalisation

Most analyses in this section compare expression values between samples (survival analysis, gene correlation, expression comparisons, GSVA, MCPcounter, Compare cohorts). Some rank genes within each sample (ssGSEA and xCell). Values normalised for both sequencing depth and gene length, such as TPM or FPKM, are valid for both kinds of comparison. The log2 transformation reduces the influence of a few very highly expressed genes.

If raw counts are uploaded:

  • Samples with more sequencing reads appear to express every gene more highly. This biases the high/low split in survival analysis, gene correlations, expression comparisons, GSVA scores and MCPcounter scores.
  • Longer genes receive more reads, which distorts the within-sample ranking used by ssGSEA and xCell.
Data Recommendation
log2(TPM+1) or log2(FPKM+1) Recommended
Normalised microarray data on a log2 scale (e.g. RMA) Suitable
TPM or FPKM without log2 transformation Works, but correlations, scores and plots are dominated by highly expressed genes. Please log2-transform.
Count-based normalisation, e.g. log2(CPM+1) or DESeq2 VST Suitable for most analyses. Not normalised for gene length, so less suitable for ssGSEA and xCell.
Raw counts Not recommended

The pre-installed TCGA cohorts use log2(RSEM normalised count+1), as provided by UCSC Xena (see 10.1). "Compare cohorts" compares the values of different cohorts directly, so only cohorts processed in the same way (same unit) can be compared meaningfully.

Example

Example

10.2.2. Patient survival information

A tab-delimited table containing the information of overall survival, progression-free survival, etc (those needed for generating a Kaplan-Meier curve or survival analysis). Please follow these rules.

  • The first column must contain sample IDs and should have the header named sample (in all lowercase). All sample IDs should exactly match those used in your gene expression and clinical data.

  • All other columns must represent pairs of event data:
    One column for the event status (censoring), with binary values: 1 (event occurred) or 0 (censored). One corresponding column for the event time (in days), labelled with the same event name followed by .time.
    For Example, when you have Overall Survival (OS) data, use one column named OS for event status and use another column named OS.time for the number of days until the event or censoring. Similary, for other types of events (e.g., DSS, DFI, PFI), follow the same format. DSS and DSS.time, DFI and DFI.time, PFI and PFI.time, etc.

  • You may include other columns in the dataset that do not follow the event/time pair format. These columns will be safely ignored and will not affect the analysis.

Example

Example

10.2.3. Metadata

Please upload a tab-delimited (.tsv) table containing metadata for the samples (patients) in your cohort. This may include information such as treatment condition, gender, grade, or cancer subtype.

  • The first column must contain the sample IDs, and the header for this column must be sample (all lowercase). All sample IDs should exactly match those used in your gene expression and clinical data.

If you do not have any metadata to include, please upload a .tsv file that contains only the sample IDs in the first column with the header sample. This ensures consistency and allows the interface to process the data correctly.

Example

Example

10.2.4. Mutation data

If you have information about which genes are mutated in which patients, you can upload this as a TSV file, with one row per mutation (a sample with several mutations has several rows).

  • A column named sample (all lowercase) containing the sample IDs. The sample IDs must exactly match those in your gene expression data; otherwise, the mutations are not assigned to the samples.
  • A column named id containing the gene names.

These two columns are required and sufficient; the column order does not matter, but we recommend sample as the first column and id as the second, as in the other files. Optionally, a column named Amino_Acid_Change (e.g. p.G12C) can be added; it is used to split the mutant samples by the type of mutation in "Deconvolution > Association with mutations" (7.3). Any other columns will be ignored.

A sample is classified as "Mutant" for a gene when this file contains any row of that gene for the sample, and as "WT" otherwise (see 4. Mutation analysis for the samples used in each analysis).

When a column is missing
Missing column What happens
sample or id An error message is shown (e.g. "The mutation data file must contain a column named 'id' for gene names.") and the file is not accepted. The other uploaded files are kept; fix the file and upload it again.
Amino_Acid_Change The file is accepted, with a note that the column was not found. All the mutation analyses work, except splitting the mutant samples by the type of mutation in 7.3: when you turn on "Split the mutant group by the type of mutation", a message is shown and the switch is turned off.

If no mutation data is uploaded, the cohort can still be added (a warning is shown). The mutation analyses then show a message that no mutation data is loaded.

Example

Example

10.2.5. How to upload

Once your files are ready in the formats described above, uploading a new cohort works as follows, in the "Upload" box below the "Registered cohort" table:

  1. Upload your Gene expression and Patient survival information files (both required) in the two upload widgets provided.
  2. Upload your Metadata file (required) and, if you have one, your Mutation data file (optional) in the two upload widgets below that.
  3. Enter a "Cohort Name" (required, must be unique across all cohorts) and an optional "Description".
  4. Check the "Previews" tabs (Expression table, Survival data, Meta data, Mutation data) to confirm each file was read correctly before submitting.
  5. Click "Add a new cohort" to register the cohort.

If you want to start over, click "Reset uploaded files" to clear everything you've uploaded so far.

Formatting reminders
  • The column containing sample (patient) IDs must be named "sample", and the column containing gene names must be named "id".
  • The Cohort Name is mandatory and must be unique.
  • Avoid special characters in the Cohort Name; use only letters, numbers, underscores, and dots.

10.3. Edit or delete the cohort

Click "Reload the database" (above the "Registered cohort" table) at any time to refresh the table with the latest data from the underlying database file.

10.3.1. Editing

  1. Go to the "Registered cohort" table in the Cohort database section.
  2. Edit the table by double-clicking on the desired field.
  3. After making your changes, click the "Save changes" button. When you see the message "saved!", your edits have been successfully applied.

10.3.2. Delete

  1. Go to the "Registered cohort" table in the Cohort database section.
  2. Select the row(s) you wish to delete.
  3. Click the "Delete selected data" button.