Chapter 10 Custom TPC modules

Here, each custom module can perform one specific TPC (TCGA, PCAWG, CCLE) analysis based on the functions introduced in Chapter 5.

All Custom TPC modules in shiny app

Figure 10.1: All Custom TPC modules in shiny app

Usually, the initial step of most modules involves the check and modification of molecular datasets through the Modify datasets[opt] widget. For some molecular types of TPC databases, there are alternative datasets for users to select, which greatly enriches the analytical possibilities

  • Among the 7 types of TCGA molecules, 4 of them have the alternative datasets.
Alternative datasets of TCGA molecules

Figure 10.2: Alternative datasets of TCGA molecules

Tips: For DNA Methylation, mean value of all CpG sites under one gene will be calculated for the gene by default. In the Modify datasets[opt] widget, users can furtherly limit its CpG sites and modify the aggregation method (e.g. median).

  • Among the 5 types of PCAWG molecules, 2 of them have the alternative datasets.

  • Among the 4 types of CCLE molecules, 1 of them has the alternative datasets.

Alternative datasets of PCAWG and CCLE molecules

Figure 10.3: Alternative datasets of PCAWG and CCLE molecules

10.1 TCGA

Table 10.1: Custom modules for TCGA analysis
Database Type Module
TCGA Comparison TCGA+GTEx: Molecular Profile Distribution (Tumor VS Normal)
TCGA Comparison TCGA: Association Between Molecular Profile and Gene Mutation
TCGA Correlation TCGA: Molecule-Molecule Correlation
TCGA Correlation TCGA: Association Between Molecular Profile and Tumor Immune Infiltration
TCGA Correlation TCGA: Association Between Molecular Profile and Immune Signature
TCGA Correlation TCGA: Association Between Molecular Profile and TMB/Stemness/MSI (Radar Show)
TCGA Correlation TCGA: Association Between Molecular Profile and Pathway Score
TCGA Survival TCGA: Molecular Profile Log-rank Analysis
TCGA Survival TCGA: Molecular Profile Cox Analysis
TCGA Dimension Reduction TCGA: Dimension Reduction Distribution

10.1.1 Comparison

10.1.1.1 TCGA+GTEx: Molecular Profile Distribution (Tumor VS Normal)

Compare molecular values between tumor and normal samples combing TCGA and GTEx datasets based on vis_toil_TvsN() and vis_toil_TvsN_cancer() functions.

  1. Select one molecular type and identifier. The mode for single cancer or pan-cancer comparison can be decided here.

  2. Adjust the visualization parameters. “TCGA Dataset only” means whether to add GTEx datasets as normal samples or not.

  3. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.

  4. Download the plot with size and format options.

The steps of module "TCGA+GTEx: Molecular Profile Distribution (Tumor VS Normal)"

Figure 10.4: The steps of module “TCGA+GTEx: Molecular Profile Distribution (Tumor VS Normal)”

10.1.1.2 TCGA: Association Between Molecular Profile and Gene Mutation

Compare molecular values between gene-mutant and gene-wild tumor samples for TCGA datasets based on vis_toil_Mut() and vis_toil_Mut_cancer() functions.

  1. Select one gene to decide the sample grouping based on its mutation status. It is necessary to click the bottom “Check” button to see if there are enough mutated samples.
  2. Select one molecular type and identifier. The mode for single cancer or pan-cancer comparison can be decided here.
  3. Adjust the visualization parameters.
  4. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.
  5. Download the plot with size and format options.
The steps of module "TCGA: Association Between Molecular Profile and Gene Mutation"

Figure 10.5: The steps of module “TCGA: Association Between Molecular Profile and Gene Mutation”

10.1.2 Correlation

10.1.2.1 TCGA: Molecule-Molecule Correlation

Compute and visualize the correlation between two molecules of TCGA databases based on vis_gene_cor() and vis_gene_cor_cancer() functions.

  1. Select two molecules. They can come from different molecular types.
  2. Adjust the visualization and analysis parameters. “Use All Cancer Types” means whether to directly use all TCGA tumor samples.
  3. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.
  4. Download the plot with size and format options.
The steps of module "TCGA: Molecule-Molecule Correlation"

Figure 10.6: The steps of module “TCGA: Molecule-Molecule Correlation”

10.1.2.2 TCGA: Association Between Molecular Profile and Tumor Immune Infiltration

Compute the correlation of pan-cancers between one molecule and tumor Immune Infiltration based on vis_gene_TIL_cor() function.

  1. Select one molecule and interesting immune infiltration estimations.
  2. Adjust the analysis parameter.
  3. Click the “Go!” button to perform analysis and display the plot in the right. The analyzed data will be also showed in right-bottom area and can be saved as CSV file.
  4. Download the plot with size and format options.
The steps of module "TCGA: Association Between Molecular Profile and Tumor Immune Infiltration"

Figure 10.7: The steps of module “TCGA: Association Between Molecular Profile and Tumor Immune Infiltration”

10.1.2.3 TCGA: Association Between Molecular Profile and Immune Signature

Compute the correlation of pan-cancers between one molecule and tumor Immune Signature based on vis_gene_immune_cor() function.

  1. Select one molecule and one source of immune signatures.
  2. Adjust the analysis parameter.
  3. Click the “Go!” button to perform analysis and display the plot in the right. The analyzed data will be also showed in right-bottom area and can be saved as CSV file.
  4. Download the plot with size and format options.
The steps of module "TCGA: Association Between Molecular Profile and Immune Signature"

Figure 10.8: The steps of module “TCGA: Association Between Molecular Profile and Immune Signature”

10.1.2.4 TCGA: Association Between Molecular Profile and TMB/Stemness/MSI (Radar Show)

Compute the correlation of pan-cancers between one molecule and TMB/Stemness/MSI index based on vis_gene_tmb_cor(), vis_gene_msi_cor() and vis_gene_stemness_cor() functions.

  1. Select one molecule and one of TMB/Stemness/MSI index.
  2. Adjust the analysis parameter.
  3. Click the “Go!” button to perform analysis and display the plot in the right. The analyzed data will be also showed in right-bottom area and can be saved as CSV file.
  4. Download the plot with size and format options.
The steps of module "TCGA: Association Between Molecular Profile and TMB/Stemness/MSI (Radar Show)"

Figure 10.9: The steps of module “TCGA: Association Between Molecular Profile and TMB/Stemness/MSI (Radar Show)”

10.1.2.5 TCGA: Association Between Molecular Profile and Pathway Score

Compute the correlation of pan-cancers between one molecule and Pathway Score based on vis_gene_pw_cor() function.

  1. Select one molecule and one pathway.
  2. Adjust the visualization and analysis parameter. “Use All Cancer Types” means whether use tumor samples from all TCGA or just one cancer.
  3. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.
  4. Download the plot with size and format options.
The steps of module "TCGA: Association Between Molecular Profile and Pathway Score"

Figure 10.10: The steps of module “TCGA: Association Between Molecular Profile and Pathway Score”

10.1.3 Survival analysis

10.1.3.1 TCGA: Molecular Profile Log-rank Analysis

Perform one molecular Log-rank survival analysis in one TCGA cancer based on tcga_surv_plot() function.

  1. Select one TCGA cancer and one molecule.
  2. Filter samples according to Age/Sex/Stage.
  3. Select survival endpoint type and molecular grouping mode.
  4. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.
  5. Download the plot with size and format options.
The steps of module "TCGA: Molecular Profile Log-rank Analysis"

Figure 10.11: The steps of module “TCGA: Molecular Profile Log-rank Analysis”

10.1.3.2 TCGA: Molecular Profile Cox Analysis

Perform one molecular Log-rank survival analysis in TCGA pan-cancers based on vis_unicox_tree() function.

  1. Select one molecule.
  2. Select survival endpoint type and molecular grouping threshold.
  3. Modify visualization parameters.
  4. Click the “Go!” button to perform analysis and display the plot in the right. The analyzed data will be also showed in right-bottom area and can be saved as CSV file.
  5. Download the plot with size and format options.
The steps of module "TCGA: Molecular Profile Cox Analysis"

Figure 10.12: The steps of module “TCGA: Molecular Profile Cox Analysis”

10.1.4 Dimension Reduction

10.1.4.1 TCGA: Dimension Reduction Distribution

Perform molecular dimension reduction analysis in one TCGA cancer based on vis_dim_dist() function.

  1. Select multiple molecules and click “Cache data” to load the molecular data. There are 3 ways to select multiple molecules.
  • “Select”: One-by-one selection;
  • “Pathway”: batch selection under one pathway;
  • “File”: Upload of identifier file.
  1. Select one cancer (or more cancers) and grouping one phenotype. Notably, user can also upload custom group.
  2. Modify analysis and visualization parameters. Three DR methods including PCA, UMAP and tSNE are supported.
  3. Click the “Go!” button to perform analysis and display the plot in the right. The analyzed data will be also showed in right-bottom area and can be saved as CSV file.
  4. Download the plot with size and format options.
The steps of module "TCGA: Dimension Reduction Distribution"

Figure 10.13: The steps of module “TCGA: Dimension Reduction Distribution”

10.2 PCAWG

Table 10.2: Custom modules for PCAWG analysis
Database Type Module
PCAWG Comparison PCAWG: Molecular Profile Distribution Across Cancer Types (Tumor VS Normal)
PCAWG Correlation PCAWG: Molecule-Molecule Correlation
PCAWG Survival PCAWG: Molecular Profile Log-rank Analysis
PCAWG Survival PCAWG: Molecular Profile Cox Analysis

10.2.1 Comparison analysis

10.2.1.1 PCAWG: Molecular Profile Distribution Across Cancer Types (Tumor VS Normal)

Compare molecular values between tumor and normal samples of PCAWG projects based on vis_pcawg_dist() function.

  1. Select one molecular type and identifier.

  2. Adjust the visualization parameters.

  3. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.

  4. Download the plot with size and format options.

The steps of module "PCAWG: Molecular Profile Distribution Across Cancer Types (Tumor VS Normal)"

Figure 10.14: The steps of module “PCAWG: Molecular Profile Distribution Across Cancer Types (Tumor VS Normal)”

10.2.2 Correlation analysis

10.2.2.1 PCAWG: Molecule-Molecule Correlation

Compute and visualize the correlation between two molecules based on PCAWG database based on vis_pcawg_gene_cor() function.

  1. Select two molecules. They can come from different datasets.
  2. Adjust the visualization and analysis parameters. “Use All Cancer Types” means whether directly use tumor samples from all PCAWG projects.
  3. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.
  4. Download the plot with size and format options.
The steps of module "PCAWG: Molecule-Molecule Correlation"

Figure 10.15: The steps of module “PCAWG: Molecule-Molecule Correlation”

10.2.3 Survival analysis

Notes: Only OS endpoint for PCAWG samples.

10.2.3.1 PCAWG: Molecular Profile Log-rank Analysis

Perform one molecular Log-rank survival analysis in one TCGA cancer.

  1. Select one PCAWG project and one molecule.
  2. Filter samples according to Age/Sex.
  3. Select molecular grouping mode.
  4. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.
  5. Download the plot with size and format options.
The steps of module "PCAWG: Molecular Profile Log-rank Analysis"

Figure 10.16: The steps of module “PCAWG: Molecular Profile Log-rank Analysis”

10.2.3.2 PCAWG: Molecular Profile Cox Analysis

Perform one molecular Log-rank survival analysis in All PCAWG projects based on vis_pcawg_unicox_tree() function.

  1. Select one molecule.
  2. Select molecular grouping threshold.
  3. Modify visualization parameters.
  4. Click the “Go!” button to perform analysis and display the plot in the right. The analyzed data will be also showed in right-bottom area and can be saved as CSV file.
  5. Download the plot with size and format options.
The steps of module "PCAWG: Molecular Profile Log-rank Analysis"

Figure 10.17: The steps of module “PCAWG: Molecular Profile Log-rank Analysis”

10.3 CCLE

Table 10.3: Custom modules for CCLE analysis
Database Type Module
CCLE Comparison CCLE: Molecular Profile Distribution Across Cancer Primary Sites
CCLE Correlation CCLE: Molecule-Molecule Correlation

10.3.1 Comparison analysis

10.3.1.1 CCLE: Molecular Profile Distribution Across Cancer Primary Sites

Compare molecular values of cancer cell lines from different primary site based on vis_ccle_tpm() function.

  1. Select one molecular type and identifier.
  2. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.
  3. Download the plot with size and format options.
The steps of module "CCLE: Molecular Profile Distribution Across Cancer Primary Sites"

Figure 10.18: The steps of module “CCLE: Molecular Profile Distribution Across Cancer Primary Sites”

10.3.2 Correlation analysis

10.3.2.1 CCLE: Molecule-Molecule Correlation

Compute and visualize the correlation between two molecules based on CCLE database based on vis_ccle_gene_cor() function.

  1. Select two molecules. They can come from different datasets.
  2. Adjust the visualization and analysis parameters. “Use All Primary Sites” means whether directly use cancer cell lines.
  3. Click the “Go!” button to perform analysis and display the plot in the right. The raw data will be also showed in right-bottom area and can be saved as CSV file.
  4. Download the plot with size and format options.
The steps of module "CCLE: Molecule-Molecule Correlation"

Figure 10.19: The steps of module “CCLE: Molecule-Molecule Correlation”