Identifications of the potential in-silico biomarkers in lung cancer tissue microbiomes.

It is postulated that the tumor tissue microbiome is one of the enabling characteristics that can either promote or suppress the ability of tumors to acquire certain hallmarks of cancer. This underscores its critical importance in carcinogenesis, cancer progression, and therapy responses. However, characterizing the tumor microbiomes is extremely challenging because of their low biomass and severe difficulties in controlling laboratory-borne contaminants, which is further aggravated by lack of comprehensively effective computational approaches to identify unique or enriched microbial species associated with cancers. Here we take advantage of a recent computational framework by Ma (2024), termed metagenome comparison (MC) framework (MCF), which can detect treatment-specific, unique or enriched OMUs (operational metagenomic unit), or US/ES (unique/enriched species) when adapted for this study. We apply the MCF to reanalyze four lung cancer tissue microbiome datasets, which include samples from Lung Adenocarcinoma (LUAD), Lung Squamous Cell Carcinoma (LUSC), and their adjacent normal tissue (NT) controls. Our analysis is structured around three distinct schemes: Scheme I-separately detecting the US/ES for each of the four lung cancer microbiome datasets; Scheme II-consolidation of the four datasets followed by detection of US/ES in the combined datasets; Scheme III-construction of the union and intersection sets of US/ES derived from the results of the preceding two schemes. The generated lists of US/ES, including enriched microbial phyla, likely hold significant biomedical value for developing diagnostic and prognostic biomarkers for lung cancer risk assessment, improving the efficacy of immunotherapy, and designing novel microbiome-based therapies in lung cancer research.