This invention belongs to the fields of
bioinformatics and
medical testing technology, specifically relating to a combination of biomarkers related to the postoperative prognosis of
colorectal cancer and their application and
prediction system. This invention, through 16S rRNA
gene sequencing of tumor mucosal samples from
colorectal cancer patients, combined with long-term postoperative follow-up data and screening, found that *
Peptostreptococcus*, *
Fusobacterium*, *
Bacteroides*, *Porphyromonas*, and *
Prevotella* are closely related to postoperative recurrence of
colorectal cancer, with an AUC of 0.81 for assessing postoperative recurrence. Based on this, this invention conducts non-
targeted metabolomics analysis, finding that combining
alanine glutamate,
putrescine,
arginine,
histidine, and
sebacic acid with the above
microbial biomarkers can improve the AUC to 0.91, significantly enhancing accuracy, specificity, and precision. This enables prognostic stratification of recurrence-free survival in colorectal
cancer patients, demonstrating higher predictive
efficacy and stability.