Immunotherapy has made great progress in hepatocellular carcinoma (HCC), yet there is still a lack of biomarkers for predicting response to it. Cancer stem cells (CSCs) are the primary cause of the tumorigenesis, metastasis, and multi-drug resistance of HCC. This study aimed to propose a novel CSCs-related cluster of HCC to predict patients' response to immunotherapy. Based on RNA-seq datasets from The Cancer Genome Atlas (TCGA) and Progenitor Cell Biology Consortium (PCBC), one-class logistic regression (OCLR) algorithm was applied to compute the stemness index (mRNAsi) of HCC patients. Unsupervised consensus clustering was performed to categorize HCC patients into two stemness subtypes which further proved to be a predictor of tumor immune microenvironment (TIME) status, immunogenomic expressions and sensitivity to neoadjuvant therapies. Finally, four machine learning algorithms (LASSO, RF, SVM-RFE and XGboost) were applied to distinguish different stemness subtypes. Thus, a five-hub-gene based classifier was constructed in TCGA and ICGC HCC datasets to predict patients' stemness subtype in a more convenient and applicable way, and this novel stemness-based classification system could facilitate the prognostic prediction and guide clinical strategies of immunotherapy and targeted therapy in HCC.

We aim to construct a hypoxia- and immune-associated risk score model to predict the prognosis of patients with pancreatic ductal adenocarcinoma (PDAC). By unsupervised consensus clustering algorithms, we generate two different hypoxia clusters. Then, we screened out 682 hypoxia-associated and 528 immune-associated PDAC differentially expressed genes (DEGs) of PDAC using Pearson correlation analysis based on the Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression project (GTEx) dataset. Seven hypoxia and immune-associated signature genes (S100A16, PPP3CA, SEMA3C, PLAU, IL18, GDF11, and NR0B1) were identified to construct a risk score model using the Univariate Cox regression and the Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression, which stratified patients into high- and low-risk groups and were further validated in the GEO and ICGC cohort. Patients in the low-risk group showed superior overall survival (OS) to their high-risk counterparts (p < 0.05). Moreover, it was suggested by multivariate Cox regression that our constructed hypoxia-associated and immune-associated prognosis signature might be used as the independent factor for prognosis prediction (p < 0.001). By CIBERSORT and ESTIMATE algorithms, we discovered that patients in high-risk groups had lower immune score, stromal score, and immune checkpoint expression such as PD-L1, and different immunocyte infiltration states compared with those low-risk patients. The mutation spectrum also differs between high- and low-risk groups. To sum up, our hypoxia- and immune-associated prognostic signature can be used as an approach to stratify the risk of PDAC.

Alzheimer’s disease, the most common form of dementia in the elderly, is a kind of neurodegenerative disease. However, its pathogenesis and diagnosis remain unclear. M6A is related to nervous system development and neurodegenerative diseases. Here in this study, using multiple RNA-seq datasets of Alzheimer’s brain tissues, along with bioinformatic analysis, we innovatively found that m6A reader protein IGF2BP2 was abnormally highly expressed in Alzheimer’s patients. After compared between Alzheimer’s and normal brain samples, and between IGF2BP2- high and IGF2BP2- low subgroups of Alzheimer’s patients, we took the shared differentially expressed genes as the relevant gene sets of IGF2PB2 affecting Alzheimer’s disease occurrence for subsequent analysis. Then, weight gene correlation analysis was conducted and 17 functional modules were identified. The module that most positively correlated with Alzheimer’s disease and IGF2PB2-high subgroups were mainly participated in ECM receptor interaction, focal adhesion, cytokine-cytokine receptor interaction, and TGF-beta signaling pathway. Afterwards, a hub gene-based model including 20 genes was constructed by LASSO regression and validated by ROC curve for Alzheimer diagnosis. Finally, we preliminarily elucidated that IGF2BP2 could bind with mRNAs in a m6A-dependent manner. This study first elucidates the pathogenic role of IGF2BP2 in Alzheimer’s disease. IGF2BP2 and its relevant m6A modifications are potential to be new diagnostic and therapeutic targets for Alzheimer’s patients.

Pancreatic cancer (PC) is a malignant digestive system tumor with a very poor prognosis. N6-methyladenosine (m6A) is mediated by a variety of readers and participates in important regulatory roles in PC. Based on TCGA_PAAD, ICGC_AU_PAAD, ICGC_CA_PAAD, GSE28735 and GSE62452 datasets, We mapped the multi-omics changes of m6A readers in PC and found that m6A readers, especially IGF2BP family genes, had specific changes and were significantly associated with poor prognosis. An unsupervised consensus clustering algorithm was used to explore the correlation between specific expression patterns of m6A readers in PC and enrichment pathways, tumor immunity and clinical molecular subtypes. Then, the principal component analysis (PCA) algorithm was used to quantify specific expression patterns and screen core genes. Machine learning algorithms such as Bootstrapping and RSF were used to quantify the expression patterns of core genes and construct a prognostic scoring model for PC patients. What's more, pharmacogenomic databases were used to screen sensitive drug targets and small molecule compounds for high-risk PC patients in an all-around and multi-angle way. Our study has not only provided new insights into personalized prognostication approaches, but also thrown light on integrating tailored risk stratification with precision therapy based on IGF2BP2-mediated m6A modification patterns.

Epithelial-mesenchymal transition (EMT) can promote carcinoma progression by multiple mechanisms; many studies demonstrated the invasiveness of pancreatic adenocarcinoma (PAAD) associated with the EMT, but how it acts through an lncRNA-dependent manner is unknown. Here, we investigated 146 samples from The Cancer Genome Atlas (TCGA) and 92 samples from the International Cancer Genome Consortium (ICGC). By gene set variation analysis (GSVA) and weighted correlation network analysis (WGCNA), we explored the EMT-related long noncoding RNAs (EMTlnc). Then, we performed univariate Cox regression analysis to screen their prognostic value for PAAD. The least absolute contraction and selection operator (LASSO) Cox regression was used to establish EMT-related lncRNA prognostic signal (EMT-LPS). In addition, we established a competitive endogenous ceRNA network. Then, we identified 33 prognostic EMTlnc as prognostic lncRNAs and established an EMT-LPS which showed strong prognostic ability in stratification analysis. By corresponding risk scores, patients were divided into low-risk and high-risk subgroups. Principal component analysis (PCA) showed that these subgroups had individual EMT status. Enrichment analysis showed that in the high-risk subgroup, biological processes, pathways, and hallmarks related to malignant tumors are more common. What is more, we constructed a nomogram that had powerful ability to predict the overall survival rate (OS) of PAAD patients in two datasets. So, EMT-LPS are a principal element in PAAD’s carcinoma progression and may help us in choosing the way of prognosis assessment and provide some clues to design the new drugs for PAAD.