AUTHOR=Yang Xia , An Huimin , He Yongtao , Fu Guoxiang , Jiang Zhinong 

TITLE=Comprehensive analysis of microbiota signature across 32 cancer types

JOURNAL=Frontiers in Oncology

VOLUME=Volume 13 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2023.1127225

DOI=10.3389/fonc.2023.1127225

ISSN=2234-943X

ABSTRACT=Microbial communities are massively resident in the human body, yet microbial dysbiosis has been reported to correlate with the tumorigenesis and progression in various cancers. However, most studies focus on specific tissue/organ, while the component signature of intratumor microbiota across various cancer types remain unclear. Based on the public databases of Bacteria in Cancer (BIC) and The Cancer Genome Atlas (TCGA), we investigated the correlation between intratumor microbial signature and clinical/genomic/immune characteristics across 32 cancer types. The relative abundance of microbial taxa in tumors compared to their paired normal tissues were observed particularly noticeable across 32 cancer types. Then, three microbiome-based clusters were determined by non-negative matrix factorization (NMF) clustering analysis based on candidate prognostic microbes, which chosen from the univariate cox regression analyses. Next, a microbiome-based scoring system (MS) were established, which composed of 64 microbes identified by the least absolute shrinkage and selection operator (LASSO) regression algorithm. In combination with mRNA sequencing data from TCGA cohort, systematically analyses were performed to evaluate the correlation between MS with survival outcomes, genomic features and immune profiles. Taken together, this study reveals the diversity and complexity of microbiome in tumor. Classified cancer into different subtypes based on intratumor microbial signatures might reasonable to reflect genomic characteristics, immune features, and survival status.