Identifying a Novel Endoplasmic Reticulum-Related Prognostic Model for Hepatocellular Carcinomas

Abstract

From the standpoint of the ER (endoplasmic reticulum), we were interested in identifying hub genes that impact clinical prognosis for HCC (hepatocellular carcinoma) patients and developing an ER-related prognostic model. Using TCGA-LIHC (The Cancer Genome Atlas-Liver Hepatocellular Carcinoma) and GSE14520 datasets, we conducted a series of analyses, which included differential gene screening, clinical prognostic analysis, Lasso regression, nomogram prediction, tumour clustering, gene functional enrichment, and tumour infiltration of immune cells. Following our screening for ER-related genes ( n = 1975 ), we conducted a Lasso regression model to obtain five hub genes, KPNA2, FMO3, SPP1, KIF2C, and LPCAT1, using TCGA-LIHC as a training set. According to risk scores, HCC samples within either the TCGG-LIHC or GSE14520 cohort were categorized into high- and low-risk groups. Compared to the high-risk group of HCC patients, patients in the low-risk group had a better prognosis of OS (overall survival) or RFS (relapse-free survival). For TCGA-LIHC training set, with the factors of risk score, stage, age, and sex, we plotted a nomogram for 1-, 3-, and 5-year survival predictions. Our model demonstrated better clinical validity in both TCGA-LIHC and GSE14520 cohorts. Additionally, events related to biological enzyme activity, biological metabolic processes, or the cell cycle were associated with the prognostic risk of ER. Furthermore, two HCC prognosis-associated tumour clusters were identified by ER hub gene-based consensus clustering. Our findings indicated a link between ER prognostic signature-related high/low risk and tumour infiltration levels of several immune cells, such as “macrophages M2/M0” and “regulatory T cells (Tregs).” Overall, we developed a novel ER-related clinical prognostic model for HCC patients.

Integration of single-cell RNA sequencing and bulk RNA sequencing to reveal an immunogenic cell death-related 5-gene panel as a prognostic model for osteosarcoma

“…From 212 OS samples, the expression patterns and clinical data of 17 ICD-related genes were derived. Next, these ICD-related genes were clustered using ConsensusClusterPlus (parameters: reps = 50, pItem = 0.8, pFeature = 1, clusterAlg = “km”, distance=“euclidean”) ( 28 – 30 ). The km and euclidean distances were used as a clustering algorithm and distance measure, respectively.…”

Construction and validation of a prognostic signature based on seven endoplasmic reticulum stress-related lncRNAs for patients with head and neck squamous cell carcinoma

“…Subsequently, we developed a novel prognostic model consisting of seven ERS-related lncRNAs. Previous studies have also constructed ERS-related models in various tumors, such as lung adenocarcinoma, breast cancer, hepatocellular carcinoma, and ovarian cancer [37][38][39][40] . However, the prognostic potential of ERS-related lncRNAs in HNSCC remains largely unexplored.…”

Chromosome instability‐associated prognostic signature and cluster investigation for cutaneous melanoma cases

“…Based on the results of the survival analyses, we utilised the ‘VennDiagram’ and ‘glmnet’ R packages to obtain five common CIN‐related prognostic genes and selected the TCGA‐SKCM as the training dataset of Lasso regression modelling. Based the yielded risk score and gene coefficient information, high‐risk and low‐risk melanoma cases were categorised, as reported previously [ 17 ]. The expression heat map of four CIN signatures ( ELAVL1 , FOXM1 , MTCH2 , MYB ) between the two groups of risk was generated by the ‘pheatmap’ R package.…”