Objective: The mortality rate of critically ill patients in ICUs is relatively high. In order to evaluate patients’ mortality risk, different scoring systems are used to help clinicians assess prognosis in ICUs, such as the Acute Physiology and Chronic Health Evaluation III (APACHE III) and the Logistic Organ Dysfunction Score (LODS). In this research, we aimed to establish and compare multiple machine learning models with physiology subscores of APACHE III—namely, the Acute Physiology Score III (APS III)—and LODS scoring systems in order to obtain better performance for ICU mortality prediction. Methods: A total number of 67,748 patients from the Medical Information Database for Intensive Care (MIMIC-IV) were enrolled, including 7055 deceased patients, and the same number of surviving patients were selected by the random downsampling technique, for a total of 14,110 patients included in the study. The enrolled patients were randomly divided into a training dataset (n = 9877) and a validation dataset (n = 4233). Fivefold cross-validation and grid search procedures were used to find and evaluate the best hyperparameters in different machine learning models. Taking the subscores of LODS and the physiology subscores that are part of the APACHE III scoring systems as input variables, four machine learning methods of XGBoost, logistic regression, support vector machine, and decision tree were used to establish ICU mortality prediction models, with AUCs as metrics. AUCs, specificity, sensitivity, positive predictive value, negative predictive value, and calibration curves were used to find the best model. Results: For the prediction of mortality risk in ICU patients, the AUC of the XGBoost model was 0.918 (95%CI, 0.915–0.922), and the AUCs of logistic regression, SVM, and decision tree were 0.872 (95%CI, 0.867–0.877), 0.872 (95%CI, 0.867–0.877), and 0.852 (95%CI, 0.847–0.857), respectively. The calibration curves of logistic regression and support vector machine performed better than the other two models in the ranges 0–40% and 70%–100%, respectively, while XGBoost performed better in the range of 40–70%. Conclusions: The mortality risk of ICU patients can be better predicted by the characteristics of the Acute Physiology Score III and the Logistic Organ Dysfunction Score with XGBoost in terms of ROC curve, sensitivity, and specificity. The XGBoost model could assist clinicians in judging in-hospital outcome of critically ill patients, especially in patients with a more uncertain survival outcome.

Cuproptosis is a new form of copper-dependent programmed cell death commonly occurring within the body. There is emerging evidence indicating that cuproptosis has a significant regulatory function in the onset and progression of cancer. However, it is still unclear how cuproptosis regulates cancer and whether other genes are involved in the regulation. Using the TCGA-COAD dataset of 512 samples, we found that seven of ten cuproptosis markers showed prognostic value in colorectal cancer (CRC) using Kaplan–Meier survival analysis. Furthermore, 31 prognostic cuproptosis-related genes were identified using weighted gene co-expression network analysis and univariate Cox analysis. Subsequently, we constructed a 7-PCRG signature using least absolute shrinkage and selection operator (LASSO)–Cox regression analysis. The risk score predicting survival in patients with CRC was evaluated. Two risk groups were classified based on their risk scores. The two groups revealed a significant difference in immune cells, such as B and T cells. Furthermore, we identified differences in many immune functions and checkpoints, including CD276 and CD28. In vitro experiments showed that a hub cuproptosis-related gene, TIGD1, could significantly regulate cuproptosis in CRC after exposure to elesclomol. This study validated that cuproptosis was closely related to the progression of CRC. Seven new cuproptosis-related genes were identified, and the function of TIGD1 in cuproptosis was preliminarily understood. Since a certain concentration of copper in CRC cells is important, cuproptosis may provide a new target for cancer therapy. This study may provide novel insights into the treatment of CRC.

MicroRNAs (miRNAs) and their target genes have been shown to play an important role in gastric cancer but have not been fully clarified. Therefore, our goal was to identify the key miRNA–mRNA regulatory network in gastric cancer by utilizing a variety of bioinformatics analyses and experiments. A total of 242 miRNAs and 1080 genes were screened from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO), respectively. Then, survival-related differentially expressed miRNAs and their differentially expressed target genes were screened. Twenty hub genes were identified from their protein–protein interaction network. After weighted gene co-expression network analysis was conducted, we selected miR-137-3p and its target gene, COL5A1, for further research. We found that miR-137-3p was significantly downregulated and that overexpression of miR-137-3p suppressed the proliferation, invasion, and migration of gastric cancer cells. Furthermore, we found that its target gene, COL5A1, could regulate the expression of another hub gene, FSTL1, by sponging miR-137-3p, which was confirmed by dual-luciferase reporter assays. Knockdown of COL5A1 inhibited the proliferation, invasion, and migration of gastric cancer cells, which could be rescued by the miR-137-3p inhibitor or overexpression of FSTL1. Ultimately, bioinformatics analyses showed that the expression of FSTL1 was highly correlated with immune infiltration.

Background: The importance of amino acid metabolism in multiple cancers was investigated by accumulating researches. But the role of amino acid metabolism-related genes (AAMRGs) played in the colorectal cancer (CRC) progression remains unclear. Methods: The clinical information and RNA sequencing of CRC were acquired from The Cancer Genome Atlas (TCGA) databases and the Gene Expression Omnibus (GEO) databases, and amino acid metabolic gene data were downloaded from a published article in the journal Cell Reports. The Cox-LASSO analysis was adopted to establish a AAMRG prognostic signature. Kaplan–Meier (K-M) survival curve and receiver operating characteristic (ROC) curve were adopted to estimate the prognostic capacity of our risk signature. RT‒qPCR was adopted to estimate the expression of AAMRGs in clinical samples. Moreover, gene set enrichment analysis (GSEA) was adopted and according to result of GSEA, further investigation was conducted in immune infiltration, somatic mutation, drug sensitivity and EMT. Result: The activity of amino acid metabolism was significantly increased in CRC. A 10-AAMRG prognostic signature was established and the CRC samples were classified into two groups (high-risk and low-risk). Risk score was analyzed to be an independent factor of CRC to affect cancer progression. The result of RT-qPCR showed that among 10 AAMRGs, the expression abundances of MRPS23 and TRAP1 in CRC tissues and adjacent normal colorectal tissues were both upregulated. According to GSEA outcomes and further analyses, significant differences were observed in immune infiltration and EMT between two risk groups. Furthermore, analysis of drug sensitivity illustrated that some chemotherapy drugs had higher IC50 values in low-risk group, including bexarotene, bicalutamide and imatinib. Conclusion: The 10-AAMRG signature we built is a prospective biomarker, which has great potentiality to predict CRC therapeutic responses and clinical prognosis.