No abstract

No abstract

Background Hepatocellular carcinoma (HCC) is a major cause of cancer death in the world. The aim of this study was to establish a new model to predict the prognosis of HCC. Materials and Methods The mRNA, miRNA and lncRNA expression profiles of early (stage I–II) and late (stage III–IV) stage HCC patients were acquired from The Cancer Genome Atlas (TCGA) database. The differentially expressed mRNAs (DEmRNAs), miRNAs (DEmiRNAs) and lncRNAs (DElncRNAs) were identified between early and late stage HCC. Key molecules associated with the prognosis, and important immune cell types in HCC were identified. The nomogram based on incorporating age, gender, stage, and all important factors was constructed to predict the survival of HCC. Results A total of 1516 DEmRNAs, 97 DEmiRNAs and 87 DElncRNAs were identified. A DElncRNA-DEmiRNA-DEmRNA regulatory network including 78 mRNAs, 50 miRNAs and 1 lncRNA was established. Among the regulatory network, 11 molecules were significantly correlated with the prognosis of HCC based on Lasso regression analysis. Then, Preadipocytes and 3 survival-associated DEmRNAs were identified as crucial biomarkers. Subsequently, a nomogram with a differentiation degree of 0.758, including 1 immune cell, 11 mRNAs and 3 miRNAs, was generated. Conclusion Our study constructed a model by incorporating clinical information, significant biomarkers and immune cells to predict the survival of HCC, which achieved a good performance.

Purpose Breast cancer (BC) is the most popular malignancy. IFN-γ plays an important role in cancer immunity. This study is to predict IFN-γ expression in BC by digital pathology. Procedures: RNA-Seq data, pathological images, and clinical data from TCGA-BRCA were retrieved. Images were segmented by the OTSU algorithm and pathomic features were extracted via PyRadiomics. The cohort was randomly divided into the training and validation set as 7:3 ratios. The maximum relevance and minimum redundancy and recursive feature elimination algorithm were used to select the features. Then the prediction model was built by the Gradient Boosting Machine algorithm. Results IFN-γ was significantly up-regulated in BC (P < 0.001), and was an independent prognostic factor for OS (HR = 0.463, 95% CI 0.316 − 0.68). The IFN-γ prediction model consisted of 6 pathomic features. The AUC of the ROC was 0.757 and 0.709 in training and validation sets, respectively. The calibration curves did not deviate from the reference line. The calculated pathomic score was an independent prognostic factor for the overall survival of BC patients (HR = 0.604, 95% CI 0.392 − 0.93). Pathomic features were found to be significantly correlated with T cells CD4 memory, T cells follicular helper and Macrophages M1 infiltrate abundance, and with immune genes, such as CD28, CD27, PDCD1 and CTLA4, (p < 0.05). Conclusions The pathomic model could predict the expression of IFN-γ and patients’ outcomes in breast cancer. The features were significantly related to the abundance of immune cell infiltration and genes.