DriverDBv4: A database for human cancer driver gene research




Gene Overview Heatmap

The Gene Overview Heatmap provides an integrated overview of various molecular features of the selected gene across multiple cancer types.
The left panel displays a heatmap based on TCGA datasets, where asterisks (*) indicate driver genes identified by multi-omics integration tools. The right panel presents corresponding results from non-TCGA datasets.



  • Differential expression (DE): Indicates whether the gene is significantly differentially expressed (p < 0.05). Red represents upregulated genes (log₂FC > 1) and green represents downregulated genes (log₂FC < –1); darker shades indicate greater fold change.
  • Mutation: Represents the number of computational tools that identify this gene as a mutation driver. The blue color becomes darker as more tools consistently identify the gene.
  • Copy Number Variation (CNV): Shows whether the gene exhibits copy number gain or loss. Red represents a gain (1) and green represents a loss (–1).
  • Methylation: Indicates the methylation state of the gene. Red represents hypermethylation (1) and green represents hypomethylation (–1).
  • Survival: Reflects whether the gene is significantly associated with patient survival (log-rank p < 0.05). Red represents oncogene-like behavior (log₂HR > 0) and green represents tumor suppressor-like behavior (log₂HR < 0); darker shades indicate stronger survival association.
  • miRNA: Displays the number of miRNAs predicted or validated to interact with the gene. The orange color becomes darker as the number of interacting miRNAs increases..

The Gene Expression interface visualizes transcript abundance (TPM) of a selected gene across multiple cancer types.
Expression distributions are grouped by sample type, mutation class, or tumor stage, presented in pan-cancer and cancer-specific views.
Switch between the tabs below to explore results under different grouping conditions.


Pan-cancer view: Expression by Sample Type

Boxplots display the TPM of the selected gene across all cancer types (x-axis), grouped and colored by sample type.
Use the legend to toggle sample types (e.g., NT, TP, TM, TR, and others). Hover over dots to view sample-level details. Hover near the box areas to view summary statistics (max, upper fence, Q3, median, Q1, min).



Cancer-Specific View: Expression by Sample Type

Cancer Type

Visualization



Pan-Cancer View: Expression Comparison (Normal vs. Tumor)

Boxplots display TPM of the selected gene across cancers, split into Normal Tissue (NT) and Primary Tumor (TP) groups.
Use the legend to toggle NT/TP groups. Hover over dots to view Cancer Type, TPM, and Sample Type, or near box areas for summary statistics.



Cancer-Specific View: Expression Comparison (Normal vs. Tumor)

Cancer Type

Visualization



Pan-Cancer View: Expression by Mutation Class

Boxplots display TPM expression of the selected gene across all cancer types (x-axis), grouped by mutation class.
Use the legend to toggle mutation impact groups (High, Moderate, Low, Modifier, Normal Tissue, Tumors without Mutation).
Hover over dots for Cancer Type, Mutation Class, and TPM, or near box areas to view summary statistics.



Cancer-Specific View: Expression by Mutation Class

Cancer Type

Visualization



Pan-Cancer View: Expression by Tumor Stage

Boxplots show TPM expression of the selected gene across all cancer types, grouped by tumor stage (Stage I–IV).
Use the legend to toggle stages. Hover over dots for Cancer Type, Stage, and TPM, or near box areas to view summary statistics.



Cancer-Specific View: Expression by Tumor Stage

Cancer Type

Visualization




The Gene Mutation interface visualizes mutation patterns and statistics of the selected gene across multiple cancer types, in correspondence with its protein regions.
Users can explore three aspects — Mutation Rate, Mutation Percent, and Exon Distribution — each offering both a Pan-Cancer View and a Cancer-Specific View. Switch between the tabs below to examine different mutation-level summaries.


Pan-Cancer View: Mutation Rate Heatmap


Cancer-Specific View: Mutation Rate Bar Chart

Cancer Type

Visualization


Pan-Cancer View: Mutation Percent Heatmap


Cancer-Specific View: Mutation Percent Bar Chart

Cancer Type

Visualization


Pan-Cancer View: Exon Mutation Distribution


Cancer-Specific View: Exon Mutation Distribution

Visualization By


Cancer Type

Visualization



The Copy Number Variation interface visualizes CNV patterns of the selected gene across multiple cancer types.
Each analysis integrates results from two CNV-identification tools — iGC and DIGGIT — and provides both pan-cancer and cancer-specific visualizations, as well as a summary table of CNV details.


Pan-Cancer View: Copy Number Variation Overview

This graph shows the copy number variation (CNV) of the selected gene across multiple cancer types.
Each gene is analyzed by two CNV tools, iGC and DIGGIT, and the plot displays the sample proportion of cancers identified by one or both tools. Bars are color-coded to indicate copy-number gain, loss, or no change. Hovering over bars provides additional information such as cancer type, CNV status, and sample proportion.


Cancer-Specific View: CNV Distribution and Correlation

Cancer Type

Visualization


CNV Summary Table


The Methylation interface visualizes the DNA methylation status of the selected gene across multiple cancer types and explores how methylation relates to gene expression.
Analyses are performed using two complementary methylation tools — MethylMix and ELMER — to identify hypermethylated and hypomethylated genes. This interface provides both pan-cancer and cancer-specific visualizations, as well as a summary table of methylation details.


Pan-Cancer View: Methylation Status Overview

This graph displays the methylation status of the selected gene across multiple cancer types.
Each gene is analyzed using two tools — MethylMix and ELMER — and the plot shows the sample proportion of cancers identified as hypermethylated, hypomethylated, or unmethylated by one or both tools.
Bars are color-coded to indicate methylation direction, and hovering over them reveals additional tool-specific information.


Cancer-Specific View: Methylation Distribution and Correlation

Cancer Type

Visualization


Methylation Summary Table


The Survival interface evaluates the prognostic impact of the selected gene on patient outcomes across multiple cancer types.
It consists of two major result sections: Single Gene Survival and Survival Synergy Network.
These analyses help users understand whether a gene’s expression is associated with favorable or unfavorable survival outcomes, and whether it exhibits synergistic effects when combined with related genes.


Single Gene Survival

This section analyzes the association between the selected gene’s expression and clinical survival outcomes across cancers.
Users can choose a survival type and stratification method before viewing results.

Survival Type:

Stratify by:

Survival Table

Select a cancer type from the table to display its Kaplan-Meier plots below.


Kaplan-Meier Plots


Gene Survival Synergy Network

This network visualizes the synergistic survival relationships between the selected gene and its related genes across cancer types. Each connection represents a synergistic effect, determined by the combined hazard ratios (HRs) of the selected gene and its partner gene. Users can refine the network by gene source, synergistic effect level, and HR direction to focus on specific biological or prognostic contexts.

Synergistic Survival Comparison

This section includes all gene pairs composed of the selected gene and its related genes whose combined hazard ratios (HRs) are greater than each gene’s individual HR in both directions. Although all pairs are displayed, only those with HR fold change ≥ 1.5 (or 2) are defined as having a synergistic survival effect.
Users can filter results by HR direction (HR > 1 or HR < 1) to focus on either risk-enhancing or protective synergies.
A table lists all qualifying gene pairs across cancer types. Select a row from the table to generate corresponding Kaplan–Meier (KM) survival plots on the right.

The miRNA Interaction interface visualizes and quantifies the regulatory relationships between a user-selected gene and its associated miRNAs across multiple cancer types.
Interactions are derived from 12 prediction tools and experimentally validated datasets (miRTarBase), with optional filtering by gene source and prediction support level.


Gene-miRNA Interaction Network

The Gene–miRNA Interaction Network illustrates predicted and validated connections between the selected gene and miRNAs. Interactions are derived from 12 prediction tools or experimentally validated data in miRTarBase. Validated interactions are shown as solid lines, while predicted interactions are shown as dotted lines. Users can filter the network by Gene Source, Minimum Prediction Support (≥6, ≥8, or ≥10 tools), and whether to display only validated interactions.


Gene-miRNA Correlation Table


The Protein Expression and mRNA–Protein Regulation interface visualizes protein-level variations, their association with mRNA expression, and post-translational modifications (PTMs) across multiple cancer types.
All results are based on the user-selected gene and can be analyzed under three conditions:
1. Across Clinical Stages – protein and mRNA–protein patterns stratified by tumor stage.
2. Across Mutation Classes – grouped by mutation impact severity.
3. Across PTM Sites – focused on specific phosphorylation sites (e.g., pY1068, pY1173).


Protein Expression by Clinical Stages across Cancer Types

Boxplots display protein expression levels of the selected gene across all TCGA cancer types. Samples are grouped by clinical stage, and users can toggle individual stages using the legend. Hover over individual boxes or dots to view additional information and statistical values. Optional filters allow users to choose specific PTM sites (None, pY1068, or pY1173) to view site-specific protein patterns.


mRNA–Protein Correlation by Clinical Stages across Cancer types

The bar chart displays Spearman correlation coefficients (ρ) between mRNA (FPKM-UQ) and protein expression (PTM-specific) across cancer types, grouped by clinical stages. Toggle individual stages using the legend to view specific stages. Hover over bars to view Cancer Type, Stage, correlation coefficient, and p-value. Selecting a bar opens a scatter plot showing gene-level correlations.


Cancer-Specific: Stage-Specific Protein Expression

Cancer Type

Visualization by stage



Protein Expression by Mutation Classes across Cancer Types Grouped

Boxplots show protein expression across all TCGA cancer types, grouped by mutation impact class, and users can toggle individual impact class using the legend. Hover over individual boxes or dots to view additional information and statistical values. Optional filters allow users to choose specific PTM sites (None, pY1068, or pY1173) to view site-specific protein patterns.


mRNA–Protein Correlation by Mutation Classes across Cancer Types

The bar chart displays Spearman correlation coefficients (ρ) between mRNA (FPKM-UQ) and protein expression (PTM-specific) across cancer types, grouped by mutation impact. Toggle individual stages using the legend to view specific impact class. Hover over bars to view Cancer Type, Stage, correlation coefficient, and p-value. Selecting a bar opens a scatter plot showing gene-level correlations.


Cancer-Specific: Mutation Impact–Specific Protein Expression

Cancer Type

Visualization by mutation



mRNA–Protein Correlation by PTM Site Across Cancer Types

The By PTM Site analysis examines how specific post-translational modifications (PTMs) influence the relationship between mRNA expression (FPKM-UQ) and protein expression across cancers. The bar chart displays Spearman correlation coefficients (ρ) between mRNA and PTM-specific protein expression across cancer types, grouped by phosphorylation site.
Use the legend to toggle individual PTM sites (e.g., None, pY1068, pY1173) for targeted comparison. Hover over bars to view Cancer type, PTM site, Correlation coefficient (ρ), and p-value. Selecting a bar opens a scatter plot showing the gene-level correlation.


The Multi-Omics Integration interface summarizes how multiple omics layers—including gene expression, mutation, CNV, methylation, and protein—converge to identify potential driver genes across cancer types.
All results are based on multi-omics prediction tools and reflect the degree of support for the user-selected gene as a driver event.
This interface includes three main result sections:
1. Integrated Multi-Omics Overview
2. Omics Connectivity Network
3. List of Multi-Omics Driver Events


Integrated Multi-Omics Overview

This visualization integrates results from multiple omics analyses and prediction tools across cancer types and reveals the importance of the selected gene.


Omics Connectivity Network

The Omics Connectivity Network illustrates relationships between the selected gene, omics layers, and cancer types.
Nodes represent entities (gene, omic type, or cancer), and edges denote associations identified by integration tools.
Nodes with more connections reflect a higher degree of multi-omics influence or cross-layer integration.


Multi-Omics Driver Event Table

The table listed the detailed information and results of of multi-omics driver events.



Citation

To support the continued development of this resource, please cite the following paper if you use DriverDBv4 in your work.

Chia-Hsin Liu , Yo-Liang Lai , Pei-Chun Shen , Hsiu-Cheng Liu , Meng-Hsin Tsai , Yu-De Wang , Wen-Jen Lin , Fang-Hsin Chen , Chia-Yang Li , Shu-Chi Wang , Mien-Chie Hung , Wei-Chung Cheng (2024).

DriverDBv4: a multi-omics integration database for cancer driver gene research. Nucleic Acids Research, 52(D1), D1246–D1252.

DOI Download citation (.nbib)

Contact

Tel: +886-4-2205-3366 #7929
Address: 5F., No. 350, Sec. 1, Jingmao Rd., Beitun Dist., Taichung City 406040, Taiwan

Copyright© 2010-2025. All Rights Reserved. ©版權所有. Ver. 1.00.003未經允許請勿任意轉載、複製或做商業用途