Screening lncRNAs with diagnostic and prognostic value for human stomach adenocarcinoma based on machine learning and mRNA-lncRNA co-expression network analysis

Non-Coding RNA as Biomarkers and Their Role in the Pathogenesis of Gastric Cancer-A Narrative Review

Non-coding RNAs (ncRNAs) represent a broad family of molecules that regulate gene expression, including microRNAs, long non-coding RNAs and circular RNAs, amongst others. Dysregulated expression of ncRNAs alters gene expression, which is implicated in the pathogenesis of several malignancies and inflammatory diseases. Gastric cancer is the fifth most frequently diagnosed cancer and the fourth most common cause of cancer-related death. Studies have found that altered expression of ncRNAs may contribute to tumourigenesis through regulating proliferation, apoptosis, drug resistance and metastasis. This review describes the potential use of ncRNAs as diagnostic and prognostic biomarkers. Moreover, we discuss the involvement of ncRNAs in the pathogenesis of gastric cancer, including their interactions with the members of major signalling pathways.

Prediction Model of Ocular Metastases in Gastric Adenocarcinoma: Machine Learning-Based Development and Interpretation Study

Background: Although gastric adenocarcinoma (GA) related ocular metastasis (OM) is rare, its occurrence indicates a more severe disease. We aimed to utilize machine learning (ML) to analyze the risk factors of GA-related OM and predict its risks. Methods: This is a retrospective cohort study. The clinical data of 3532 GA patients were collected and randomly classified into training and validation sets in a ratio of 7:3. Those with or without OM were classified into OM and non-OM (NOM) groups. Univariate and multivariate logistic regression analyses and least absolute shrinkage and selection operator were conducted. We integrated the variables identified through feature importance ranking and further refined the selection process using forward sequential feature selection based on random forest (RF) algorithm before incorporating them into the ML model. We applied six ML algorithms to construct the predictive GA model. The area under the receiver operating characteristic (ROC) curve indicated the model's predictive ability. Also, we established a network risk calculator based on the best performance model. We used Shapley additive interpretation (SHAP) to identify risk factors and to confirm the interpretability of the black box model. We have de-identified all patient details. Results: The ML model, consisting of 13 variables, achieved an optimal predictive performance using the gradient boosting machine (GBM) model, with an impressive area under the curve (AUC) of 0.997 in the test set. Utilizing the SHAP method, we identified crucial factors for OM in GA patients, including LDL, CA724, CEA, AFP, CA125, Hb, CA153, and Ca2+. Additionally, we validated the model's reliability through an analysis of two patient cases and developed a functional online web prediction calculator based on the GBM model. Conclusion: We used the ML method to establish a risk prediction model for GA-related OM and showed that GBM performed best among the six ML models. The model may identify patients with GA-related OM to provide early and timely treatment.

SEC24D gene as a biomarker in human cancers and its association with CD8+ T cell immune cell infiltration

OBJECTIVE

The SEC24D (SEC24 Homolog D, COPII Coat Complex Component) gene belongs to the SEC24 subfamily of genes. The protein encoded by this gene, along with its other binding partners, mediates the transport of newly-synthesized proteins from the endoplasmic reticulum to the Golgi apparatus.

METHODS

A pan-cancer analysis of this gene, as well as its diagnostic and prognostic implications, are lacking in the medical literature. First, we analyzed SEC24D gene expression, its prognostic effect, promoter methylation level, genetic alteration landscape, pathways, CD8+ T immune cell infiltration, and gene-drug network in various types of cancer through various online databases and bioinformatic tools. Then, we performed the expression and methylation validation analysis of the SEC24D gene on cell lines using RNA sequencing (RNA-seq) and targeted bisulfite sequencing (bisulfite-seq) techniques.

RESULTS

Bioinformatic analysis showed that the SEC24D gene was overexpressed in metastasis across Kidney Renal Clear Cell Carcinoma (KIRC), Lung Squamous Cell Carcinoma (LUSC), and Stomach Adenocarcinoma (STAD) patients and was a prognostic risk factor. Then, using RNA sequencing and targeted bisulfite sequencing analysis, it was validated in cell lines that SEC24D was overexpressed and hypomethylated in KIRC patients. Mutational analysis revealed that SEC24D was mutated less frequently in KIRC, LUSC, and STAD patients. It was further observed that CD8+ T cell infiltration levels were increased in SEC24D-overexpressed KIRC, LUSC, and STAD samples. Pathway enrichment analysis of SEC24D-associated genes revealed their participation in two important pathways. Moreover, we suggested a few valuable drugs for treating KIRC, LUSC, and STAD patients with respect to overexpressed SEC24D.

CONCLUSION

This is the first pan-cancer study that details the oncogenic roles of SEC24D among different cancers.

Identification of the three subtypes and the prognostic characteristics of stomach adenocarcinoma: analysis of the hypoxia-related long non-coding RNAs

Stomach adenocarcinoma (STAD) is one of the most commonly diagnosed cancers. This study analyzed the subtypes and characteristics of STAD subtypes by analyzing hypoxia pathway-related lncRNAs. Potential hub lncRNAs were found and a prognostic model was constructed. Expression profiling data and clinical information of STAD were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). Metabolic pathway scores were calculated using single-sample gene set enrichment analysis (ssGSEA) method. Tumor immune microenvironment scores of the samples were assessed by ESTIMATE, MCP-counter, and ssGSEA. Functional analysis of lncRNAs, construction of risk models, and drug sensitivity analysis were performed. Pathway analysis revealed that the hypoxia pathway was a prognostic risk factor. Molecular subtypes were developed based on the hypoxia score-related lncRNAs. Three molecular subtypes (C1, C2, and C3) for gastric STAD were determined. The worst prognosis was in the C2, which was also characterized by the maximum hypoxia pathway-related scores and the maximum immune score. A majority of the immune checkpoints and chemokines were high-expressed in the C2 subtype. Mutations in the C2 subtype were significantly lower than the C1 and C3 subtypes. The subtypes differed in terms of functional and metabolic pathways. Eight hub indicator lncRNAs (MSC-AS1, AC037198.1, LINC00968, AL139393.3, LINC02544, BOLA3-AS1, MIR1915HG, and AC107021.2) capable of predicting patient prognosis were identified. Three hypoxia lncRNA-related molecular subtypes characterized by different prognostic and immune conditions were identified. The results maybe can provide a theoretical basis to improve the clinical diagnosis and treatment of STAD.

Machine Learning for Future Subtyping of the Tumor Microenvironment of Gastro-Esophageal Adenocarcinomas

Tumor progression involves an intricate interplay between malignant cells and their surrounding tumor microenvironment (TME) at specific sites. The TME is dynamic and is composed of stromal, parenchymal, and immune cells, which mediate cancer progression and therapy resistance. Evidence from preclinical and clinical studies revealed that TME targeting and reprogramming can be a promising approach to achieve anti-tumor effects in several cancers, including in GEA. Thus, it is of great interest to use modern technology to understand the relevant components of programming the TME. Here, we discuss the approach of machine learning, which recently gained increasing interest recently because of its ability to measure tumor parameters at the cellular level, reveal global features of relevance, and generate prognostic models. In this review, we discuss the relevant stromal composition of the TME in GEAs and discuss how they could be integrated. We also review the current progress in the application of machine learning in different medical disciplines that are relevant for the management and study of GEA.

Applying immune-related lncRNA pairs to construct a prognostic signature and predict the immune landscape of stomach adenocarcinoma

Background: Long noncoding RNAs (lncRNAs) are associated with the survival of cancer patients. We constructed an immune-related lncRNA (irlncRNA) pair signature for stomach adenocarcinoma (STAD).Research design and methods: irlncRNAs were identified via coexpression analysis with immune-related genes. Differentially expressed irlncRNAs (DEirlncRNAs) were paired. Least absolute shrinkage and selection operator (LASSO) and multivariate Cox proportional hazards regression methods were used to construct the signature. We calculated the area under the receiver operating characteristic (ROC) curve and determined the best cutoff value according to the Akaike information criterion (AIC). Patients were divided into high - and low-risk groups, and differences in immune cell infiltration, tumor mutation burden (TMB) and drug treatment effects between the groups were explored according to the risk score.Results: An 8-irlncRNA-pair signature was constructed and proven to be a strong prognosis predictor in STAD patients through external verification. Moreover, the risk score was identified as an independent prognostic factor. There were significant differences in immune cell infiltration and the response to several drug treatments between patients with high and low risk scores, and the risk score was negatively correlated with TMB.Conclusions: The signature consisting of 8 irlncRNA pairs showed good prognostic predictive value.

Identification of prognostic long non-coding RNA signature with potential drugs in hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is the primary malignancy in the liver with high rate of death and recurrence. Novel prognostic model would be crucial for early diagnosis and improved clinical decision. The study aims to provide an effective lncRNA-based signature to predict survival time and tumor recurrence for HCC. Based on public database, lncRNA-based classifiers for overall survival and tumor recurrence were built with regression analysis and cross validation strategy. According to the risk-score of the classifiers, the whole cohorts were divided into groups with high and low risk. Afterwards, the efficiency of the lncRNA-based classifiers was evaluated and compared with other clinical factors. Finally, candidate small molecules for high risk groups were further screened using drug response databases to explore potential drugs for HCC treatment.

Screening lncRNAs with diagnostic and prognostic value for human stomach adenocarcinoma based on machine learning and mRNA-lncRNA co-expression network analysis

BACKGROUND

Stomach adenocarcinoma (STAD), is one of the most lethal malignancies around the world. The aim of this study was to find the long noncoding RNAs (lncRNAs) acting as diagnostic and prognostic biomarker of STAD.

METHODS

Base on TCGA dataset, the differentially expressed mRNAs (DEmRNAs) and lncRNAs (DElncRNAs) were identified between STAD and normal tissue. The machine learning and survival analysis were performed to evaluate the potential diagnostic and prognostic value of lncRNAs for STAD. We also build the co-expression network and functional annotation. The expression of selected candidate mRNAs and lncRNAs were validated by Quantitative real-time polymerase chain reaction (qRT-PCR) and GSE27342 dataset. GSE27342 dataset were also to perform gene set enrichment analysis.

RESULTS

A total of 814 DEmRNAs and 106 DElncRNAs between STAD and normal tissue were obtained. FOXD2-AS1, LINC01235, and RP11-598F7.5 were defined as optimal diagnostic lncRNA biomarkers for STAD. The area under curve (AUC) of the decision tree model, random forests model, and support vector machine (SVM) model were 0.797, 0.981, and 0.983, and the specificity and sensitivity of the three model were 75.0% and 97.1%, 96.9% and 96%, and 96.9% and 97.1%, respectively. Among them, LINC01235 was not only an optimal diagnostic lncRNA biomarkers, but also related to survival time. The expression of three DEmRNAs (ESM1, WNT2, and COL10A1) and three optimal diagnostic lncRNAs biomarkers (FOXD2-AS1, RP11-598F7.5, and LINC01235) in qRT-PCR validation was were consistent with our integrated analysis. Except for FOXD2-AS1, ESM1, WNT2, COL10A1, and LINC01235 were upregulated in STAD, which was consistent with our integration results. Gene set enrichment analysis results indicated that DNA replication, Cell cycle, ECM-receptor interaction, and P53 signaling pathway were four significantly enriched pathways in STAD.

CONCLUSION

Our study identified three DElncRNAs as potential diagnostic biomarkers of STAD. Among them, LINC01235 also was a prognostic lncRNA biomarkers.