Investigating tryptophan metabolism in colorectal cancer using Single-cell RNA sequencing based on machine learning techniques

Background: Colorectal cancer (CRC) is characterized by genetic variation, epigenetic alterations, microenvironmental imbalance, and metabolic reprogramming. Currently, abnormalities amino acid metabolism has been shown to play an important role in the occurrence and progression of CRC. Methods: AUCell, UCell, singscore, ssGSEA and AddModuleScore algorithms were used to determine the pattern of tryptophan metabolism in CRC at the cellular level. Differential expression and correlation analyses were performed to identify core candidate genes associated with upregulation of metabolic activity. Four machine learning algorithms——random forest, Boruta, LASSO, and gradient boosting machine—were further integrated for feature selection. Finally, to enhance robustness and reduce algorithm‑specific bias, the results of these algorithms were combined to identify the key feature genes related to tryptophan metabolism in CRC. Results: The findings demonstrated significant differences in tryptophan metabolic activity among different cell types in CRC, with macrophages and Paneth cells exhibiting higher activity. Among the tryptophan metabolism-related genes, CYP1A1 and aryl hydrocarbon receptor (AHR) were significantly upregulated in CRC, suggesting their involvement in regulating of immune response and inflammatory responses. Conclusions: This study reveals, for the first time, the cellular pattern of tryptophan metabolism in CRC, with macrophages and Paneth cells playing a major role in tumor development. CYP1A1 and AHR were identified as consensus feature‑selected genes involved in tryptophan metabolism in CRC, highlighting their potential as biomarkers and therapeutic targets.