Demography-dependent variability in the human tumor mycobiome

Area of Science:

• Microbiology
• Oncology
• Bioinformatics

Background:

• The tumor mycobiome, the community of fungi within tumors, is increasingly recognized for its potential prognostic and diagnostic value in cancer.
• Patient demographics such as age, sex, body mass index (BMI), and race may influence the composition of the tumor mycobiome.
• Previous analyses of tumor microbiome data have faced critiques regarding data processing, potentially leading to false signals.

Purpose of the Study:

• To investigate how patient demographic characteristics influence the composition of the intratumor mycobiome across various cancer types.
• To address and mitigate potential biases arising from data processing techniques in mycobiome analysis.
• To identify specific fungal species whose abundance varies significantly with demographic factors within different cancers.

Main Methods:

• Explored 14 combinations of data transformation and batch correction methods on data from 224 fungal species across 13 cancer types.
• Utilized propensity scores to adjust for confounders like histological type and tumor stage.
• Identified significant fungal species abundance differences by normalizing data across all 14 processing combinations to minimize false positive outcomes.

Main Results:

• Observed significant variations in the abundance of 24 fungal species within tumors, linked to specific demographic characteristics.
• A substantial portion (20 out of 24) of these significant differences were associated with race within particular cancer types.
• Demonstrated intricate interactions between the mycobiome, cancer type, and patient demographics.

Conclusions:

• Patient race is a critical factor influencing tumor mycobiome composition and requires consideration in cancer research.
• Robust data processing techniques, including normalization and batch correction, are essential for reliable mycobiome analysis.
• The tumor mycobiome represents a promising area for precision oncology, with potential for personalized diagnostics and treatments informed by demographic factors.