Identification of microbial species and proteins associated with colorectal cancer by reanalyzing CPTAC proteomic datasets
Abstract Microbiome research has revealed associations between microbial species and colorectal cancer (CRC). Most of the existing research relied on metagenomic data. We leveraged a tool that we recently developed for detecting human and microbial peptides from (meta)proteomics data to reanalyze Cl...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-04-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-025-97984-3 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Microbiome research has revealed associations between microbial species and colorectal cancer (CRC). Most of the existing research relied on metagenomic data. We leveraged a tool that we recently developed for detecting human and microbial peptides from (meta)proteomics data to reanalyze Clinical Proteomic Tumor Analysis Consortium CRC proteomics datasets. Our analyses revealed potential microbial species and proteins that are associated with CRC, especially when analyzing multiplexed proteomics data consisting of cancerous and healthy tissue taken from the same individuals. Many of the identified proteins are associated with species with known links to CRC, such as the fungi Aspergillus kawachii, but many are unstudied or their specific roles unknown. Proteins from other microbial species, such as Paenibacillus cellulosilyticus, were also identified in the samples. We showed that Aspergillus kawachii and others are depleted overall in cancer samples, which is consistent with a previous genomic-based multi-cohort study. Our analysis also revealed that some proteins belonging to this species are more abundantly detected, while others in this and other species are not. Further, we showed that microbial identifications could be used to build predictive models for tumor detection, but caution needs to be taken when applying such models trained on one dataset to another due to the substantial impacts of different experimental techniques on peptide detection profiles. |
|---|---|
| ISSN: | 2045-2322 |