Related Experiment Video
Updated: Jun 29, 2026
08:32
Enrichment and Characterization of the Tumor Immune and Non-immune Microenvironments in Established Subcutaneous Murine Tumors
Published on: June 7, 2018
9.6K
Systematic evaluation of TCGA tumor microbiota reveals context-dependent reliability
Chenchen Ma1,2, Changxing Su1,2, Jiaxuan Li1,2
1Department of Human Cell Biology and Genetics, School of Medicine, Southern University of Science and Technology, Shenzhen, Guangdong, China.
Msystems
|April 22, 2026
Summary
This study benchmarks The Cancer Genome Atlas (TCGA) microbial profiles, revealing variable accuracy in detecting oncomicrobes and unreliable host-microbe associations. A new framework and web portal (MOMAC2) help identify trustworthy cancer microbiome signals for future research.
Area of Science:
- Oncology
- Microbiology
- Bioinformatics
Background:
- Tumor-associated bacteria are crucial in cancer biology, but the reliability of The Cancer Genome Atlas (TCGA) microbial data is uncertain.
- Existing TCGA microbial profiles (TMPs) are widely used, yet their accuracy and consistency in host-microbe association studies require systematic evaluation.
Purpose of the Study:
- To systematically benchmark TCGA microbial profiles for consistency, accuracy, and reliability in host-microbe association studies across 24 cancer types.
- To develop a statistical framework to differentiate true biological signals from spurious associations in tumor microbiome data.
- To introduce the Multi-Omics and Microbiome Associations in Cancer 2 (MOMAC2) web portal for accessing reliability-graded findings.
Main Methods:
- Systematic benchmarking of two leading TCGA microbial profiles (TMPs) focusing on bacterial components.
- Assessment of TMP accuracy for known oncomicrobes (e.g., HPV, Helicobacter pylori).
- Evaluation of host-microbe association concordance across gene expression, methylation, and protein data.
- Development of a permutation-based statistical framework to assess reliability.
- Experimental validation of high-confidence associations using co-culture models.
Main Results:
- TCGA microbial profiles showed substantial agreement in microbial composition but variable accuracy for specific oncomicrobes.
- Host-microbe association concordance was moderate for gene expression but low for methylation and protein data.
- A significant portion of associations, particularly those involving cell type and patient survival, were found to be statistically spurious.
- High-confidence associations identified HPV-driven axes and implicated Streptococcus anginosus in promoting oral cancer cell proliferation and migration.
Conclusions:
- TCGA microbial data requires careful, multi-layered validation for robust biological insights due to variable reliability.
- The developed statistical framework and MOMAC2 web portal provide essential tools for interpreting and validating tumor microbiome data.
- Validated associations, such as the role of Streptococcus anginosus, can guide experimental research and uncover novel cancer mechanisms.
