What question did this study set out to answer?

This research aims to explore the interconnection between nucleotide metabolism and mast cells in stomach adenocarcinoma to improve risk stratification and treatment strategies.

July 14, 2026Discover Oncology0 citationsOpen Access

Machine learning and multi-omic empowered risk stratification and therapeutic framework targeting nucleotide metabolism and mast cell for stomach adenocarcinoma patients

YSYue ShiChangchun University of Chinese Medicine JZJ ZhangChangchun University of Chinese Medicine XMXiaoping MenThe Affiliated Hospital to Changchun University of Chinese Medicine

Key Points

This research aims to explore the interconnection between nucleotide metabolism and mast cells in stomach adenocarcinoma to improve risk stratification and treatment strategies.
Utilized bulk profiles from GSE26899 and GSE10326 for gene identification through bioinformatics analysis using Limma, ssGSEA, and WGCNA.
Employed Lasso-cox regression for predictive model creation and identification of hub genes in TCGA-STAD and GSE84437 datasets.
Applied a deep learning algorithm for risk stratification and conducted in vitro studies for evaluating the relationship between hub genes and cancer cell behaviors.
MNM guided prognostic forecasting for STAD patients, highlighting its predictive value.
Identified UCK2 as a key hub gene influencing STAD progression based on both in silico and in vitro evidence.
Introduced HG-5-113-01 as a promising therapeutic candidate targeting UCK2 for STAD treatment.

Abstract

Abstract Background Mast cell and nucleotide metabolism(NM) encodes the Stomach adenocarcinoma (STAD) progression and tumor immune microenvironment(TME) heterogeneity. Research targeting deciphering NM-mast cell axis in STAD can pave the way for the deeper understanding of STAD pathogenesis. Methods STAD stomach tissue bulk profiles(GSE26899 and GSE10326) were utilized for identification of Mast cell and nucleotide metabolism(MNM)-associated shared differentially expressed genes(DEGs) via integrative bioinformatic analysis, including Limma, ssGSEA and WGCNA. Next, Lasso-cox regression was cross-employed for elaboration of predictive model and MNM-associated hub gene identification in TCGA-STAD and GSE84437 bulk profiles for STAD patients. Besides, deep learning algorithm(SOM) was utilized for risk stratification for STAD patients in TCGA-STAD dataset. Indeed, we also elucidated the molecular and immune heterogeneity of hub gene at bulk(TCGA-STAD cohort) and single-cell transcriptomic(GSE158937) levels of STAD patients, especially in virtual malignant cells at spatial and temporal manners. Besides, GSCA database with ridge regression were cross-performed for identification of optimal therapeutic framework for STAD patient treatment and then validated by molecular docking. Finally, in vitro study examined the relationship between hub gene and STAD cancer cell proliferation, growth and metastasis. Results MNM can guide the STAD patient prognostic forecasting and risk stratification. Particularly, UCK2 can be considered as MNM-associated hub gene involved in STAD pathogenesis proved by in silico and in vitro studies. Indeed, HG-5-113-01 should be considered as drug reproposing framework targeting UCK2 for the treatment of STAD. Conclusion Our study first discovered the integration of MNM mechanisms and predictive roles for STAD patients by combining artificial intelligence(AI) and multi-omic studies.

Ask AI

Helpful

Bookmark

View Full Paper