We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
基于单细胞 RNA-seq 和 Bulk RNA-seq 数据构建 巨噬细胞相关基因的肝癌预后风险预测模型.
- Authors
邓洁莲; 郑薇; 李康杰; 张聪; 张源; 谢彪; 钟晓妮
- Abstract
Objective The aim of this study was to identify macrophage related genes (MRGs) in liver cancer and construct a prognostic risk prediction model for liver cancer. Methods The liver cancer scRNA-seq data from the GEO database were downloaded to identify genes specifically expressed in macrophages as MRGs. The GO and KEGG functional enrichment analyses on MRGS were conducted. In the TCGALIHC dataset of the TCGA database, multiple random sampling single factor Cox regression for single-factor Cox regression, LASSO regression, and multivariate Cox regression analyses were employed to identify MRGs for liver cancer prognosis prediction, and a liver cancer prognostic prediction model was constructed. Results The results obtained 8 major cell types, including those containing macrophages through clustering using scRNA seq data from the GEO database. The proportion of macrophages in the immune microenvironment of liver cancer was significantly higher than that of normal tissues (P = 0.016), and genes such as SPP1, RNASE1, and MMP9 were highly expressed. Multiple metabolic pathways, including purine metabolism, citric acid cycle, and drug metabolism cytochrome P450 were activated in liver cancer associated macrophages. This study identified 777 MRGs from liver cancer scRNA-seq (LogFC>0.25, P<0.05), which mainly involved in functions such as actin binding and regulation of immune receptor activity. Seven MRGs, including ATP1B3, ATP6V0B, HBEGF, KLF2, NR1H3, RAB10, and SPP1 were selected from the 169 stable prognostic genes (P<0.05) for the construction of the prognosis model. The AUC values for the 1,3, and 5- year survival outcomes of the model in the TCGA liver cancer cohort were 0.791,0.791, and 0.751, respectively. In the validation ICGC cohort, they were 0. 614,0.682, and 0.688, respectively, demonstrating good predictive performance. In liver cancer patients with high prognosis risk scores, the expression of macrophages MO, neutrophils, and regulatory T cells was higher (P<0.05), and immunosuppression and immune activation coexisted. Conclusion Liver cancer MRGs can serve as potential biomarkers for predicting the prognosis of liver cancer patients. These liver cancer MRGs are mainly associated with actin binding, immune receptor activity, and infiltration of various immune cells.
- Subjects
PURINE metabolism; LIVER tumors; RISK assessment; MACROPHAGES; PREDICTION models; DESCRIPTIVE statistics; GENES; RNA; IMMUNOSUPPRESSION; DISEASE risk factors
- Publication
Practical Oncology Journal, 2023, Vol 37, Issue 5, p403
- ISSN
1002-3070
- Publication type
Article
- DOI
10.11904/j.issn.1002-3070.2023.05.003