In primary glioblastoma fewer tumor copy number segments of the F13A1 gene are associated with poorer survival

Constructing a Glioblastoma Prognostic Model Related to Fatty Acid Metabolism Using Machine Learning and Identifying F13A1 as a Potential Target

Background: Increased fatty acid metabolism (FAM) is an important marker of tumor metabolism. However, the characterization and function of FAM-related genes in glioblastoma (GBM) have not been fully explored. Method: In the TCGA-GBM cohort, FAM-related genes were divided into three clusters (C1, C2, and C3), and the DEGs between the clusters and those in the normal group and GBM cohort were considered key genes. On the basis of 10 kinds of machine learning methods, we used 101 combinations of algorithms to construct prognostic models and obtain the best model. In addition, we also validated the model in the GSE43378, GSE83300, CGGA, and REMBRANDT datasets. We also conducted a multifaceted analysis of F13A1, which plays an important role in the best model. Results: C2, with the worst prognosis, may be associated with an immunosuppressive phenotype, which may be related to positive regulation of cell adhesion and lymphocyte-mediated immunity. Using multiple machine learning methods, we identified RSF as the best prognostic model. In the RSF model, F13A1 accounts for the most important contribution. F13A1 can support GBM malignant tumor cells by promoting fatty acid metabolism in GBM macrophages, leading to a poor prognosis for patients. This metabolic reprogramming not only enhances the survival and proliferation of macrophages, but also may promote the growth, invasion, and metastasis of GBM cells by secreting growth factors and cytokines. F13A1 is significantly correlated with immune-related molecules, including IL2RA, which may activate immunity, and IL10, which suggests immune suppression. F13A1 also interferes with immune cell recognition and killing of GBM cells by affecting MHC molecules. Conclusions: The prognostic model developed here helps us to further enhance our understanding of FAM in GBM and provides a compelling avenue for the clinical prediction of patient prognosis and treatment. We also identified F13A1 as a possibly novel tumor marker for GBM which can support GBM malignant tumor cells by promoting fatty acid metabolism in GBM macrophages.

Comprehensive analyses for the coagulation and macrophage-related genes to reveal their joint roles in the prognosis and immunotherapy of lung adenocarcinoma patients

Purpose

This study aims to explore novel biomarkers related to the coagulation process and tumor-associated macrophage (TAM) infiltration in lung adenocarcinoma (LUAD).

Methods

The macrophage M2-related genes were obtained by Weighted Gene Co-expression Network Analysis (WGCNA) in bulk RNA-seq data, while the TAM marker genes were identified by analyzing the scRNA-seq data, and the coagulation-associated genes were obtained from MSigDB and KEGG databases. Survival analysis was performed for the intersectional genes. A risk score model was subsequently constructed based on the survival-related genes for prognosis prediction and validated in external datasets.

Results

In total, 33 coagulation and macrophage-related (COMAR) genes were obtained, 19 of which were selected for the risk score model construction. Finally, 10 survival-associated genes (APOE, ARRB2, C1QB, F13A1, FCGR2A, FYN, ITGB2, MMP9, OLR1, and VSIG4) were involved in the COMAR risk score model. According to the risk score, patients were equally divided into low- and high-risk groups, and the prognosis of patients in the high-risk group was significantly worse than that in the low-risk group. The ROC curve indicated that the risk score model had high sensitivity and specificity, which was validated in multiple external datasets. Moreover, the model also had high efficacy in predicting the clinical outcomes of LUAD patients who received anti-PD-1/PD-L1 immunotherapy.

Conclusion

The COMAR risk score model constructed in this study has excellent predictive value for the prognosis and immunotherapeutic clinical outcomes of patients with LUAD, which provides potential biomarkers for the treatment and prognostic prediction.

Integration of Raman spectra with transcriptome data in glioblastoma multiforme defines tumour subtypes and predicts patient outcome

Raman spectroscopy is an imaging technique that has been applied to assess molecular compositions of living cells to characterize cell types and states. However, owing to the diverse molecular species in cells and challenges of assigning peaks to specific molecules, it has not been clear how to interpret cellular Raman spectra. Here, we provide firm evidence that cellular Raman spectra (RS) and transcriptomic profiles of glioblastoma can be computationally connected and thus interpreted. We find that the dimensions of high-dimensional RS and transcriptomes can be reduced and connected linearly through a shared low-dimensional subspace. Accordingly, we were able to predict global gene expression profiles by applying the calculated transformation matrix to Raman spectra and vice versa. From these analyses, we extract a minimal gene expression signature associated with specific RS profiles and predictive of disease outcome.

Identification and characterization of N6-methyladenosine modification of circRNAs in glioblastoma

This research systematically profiled the global N6-methyladenosine modification pattern of circular RNAs (circRNAs) in glioblastoma (GBM). Based on RNA methylation sequencing (MeRIP sequencing or N6-methyladenosine sequencing) and RNA sequencing, we described the N6-methyladenosine modification status and gene expression of circRNAs in GBM and normal brain tissues. N6-methyladenosine-related circRNAs were immunoprecipitated and validated by real-time quantitative PCR. Bioinformatics analysis and related screening were carried out. Compared with those of the NC group, the circRNAs from GBM exhibited 1370 new N6-methyladenosine peaks and 1322 missing N6-methyladenosine peaks. Among the loci associated with altered N6-methyladenosine peaks, 1298 were up-regulated and 1905 were down-regulated. The N6-methyladenosine level tended to be positively correlated with circRNA expression. Bioinformatics analysis was used to predict the biological function of N6-methyladenosine-modified circRNAs and the corresponding signalling pathways. In addition, through PCR validation combined with clinical data mining, we identified five molecules of interest (BUB1, C1S, DTHD1, F13A1 and NDC80) that could be initial candidates for further study of the function and mechanism of N6-methyladenosine-mediated GBM development. In conclusion, our findings demonstrated the N6-methyladenosine modification pattern of circRNAs in human GBM, revealing the possible roles of N6-methyladenosine-mediated novel noncoding RNAs in the origin and progression of GBM.

Identification of five important genes to predict glioblastoma subtypes

BACKGROUND

Glioblastoma (GBM), the most common and aggressive primary brain tumour in adults, has been classified into three subtypes: classical, mesenchymal, and proneural. While the original classification relied on an 840 gene-set, further clarification on true GBM subtypes uses a 150-gene signature to accurately classify GBM into the three subtypes. We hypothesized whether a machine learning approach could be used to identify a smaller gene-set to accurately predict GBM subtype.

METHODS

Using a supervised machine learning approach, extreme gradient boosting (XGBoost), we developed a classifier to predict the three subtypes of glioblastoma (GBM): classical, mesenchymal, and proneural. We tested the classifier on in-house GBM tissue, cell lines, and xenograft samples to predict their subtype.

RESULTS

We identified the five most important genes for characterizing the three subtypes based on genes that often exhibited high Importance Scores in our XGBoost analyses. On average, this approach achieved 80.12% accuracy in predicting these three subtypes of GBM. Furthermore, we applied our five-gene classifier to successfully predict the subtype of GBM samples at our centre.

CONCLUSION

Our 5-gene set classifier is the smallest classifier to date that can predict GBM subtypes with high accuracy, which could facilitate the future development of a five-gene subtype diagnostic biomarker for routine assays in GBM samples.

von Willebrand Factor Gene Expression in Primary Lower Grade Glioma: Mutually Co-Occurring Mutations in von Willebrand Factor, ATRX, and TP53

BACKGROUND

Venous thromboembolism is a common complication in patients with glioma. The clotting factor von Willebrand factor (VWF) is a highly adhesive procoagulant molecule that mediates platelet adhesion to endothelial and subendothelial surfaces. In the current analysis, we examined The Cancer Genome Atlas (TCGA) data to assess the VWF gene in patients with lower grade gliomas.

METHODS

For newly diagnosed gliomas, we evaluated the association between VWF and overall survival in the Genomic Data Commons TCGA Lower Grade Glioma (LGG) dataset in TCGA. Simple statistics were calculated to identify patterns of mutual exclusivity or co-occurrence of VWF mutations. For each pair of query genes an odds ratio was calculated that indicates the likelihood that the mutations in the two genes are mutually exclusive or co-occurrent across the selected cases. To determine whether the identified relationship was significant for a gene pair, Fisher's exact test was performed.

RESULTS

Lower grade gliomas with less VWF gene expression had significantly better survival than those with more VWF gene expression (hazard ratio 0.64, 95% confidence interval 0.44 to 0.92, p=0.015 log rank test). When we analyzed the data with Cox regression, VWF expression had a significant effect on survival (p=0.02) that was unrelated to the effect of IDH1 expression (p=0.062), TP53 expression (p=0.135), independent of ATRX expression (p=0.021) and histology (astrocytoma versus oligoastrocytoma and oligodendroglioma, p=0.002). VWF mutations significantly co-occur with mutations in TP53 and ATRX (p<0.001).

CONCLUSION

The deleterious prognostic effect of VWF expression and its co-occurrent mutations with TP53 and ATRX in lower grade gliomas are not surprising, given VWF's role in other cancers. Therefore, VWF gene expression may be a clinically important risk marker in lower grade glioma.

Increased expression of von Willebrand factor gene is associated with poorer survival in primary lower grade glioma

Background

Venous thromboembolism is a common complication in patients with glioma. The clotting factor von Willebrand factor (VWF) is a highly adhesive procoagulant molecule that mediates platelet adhesion to endothelial and subendothelial surfaces. In the current analysis, we examined The Cancer Genome Atlas (TCGA) data to assess the effect of VWF gene expression on prognosis in patients with lower grade gliomas (LGGs).

Methods

For newly diagnosed gliomas, we evaluated the association between and overall survival in the genomic data commons TCGA LGG dataset in TCGA. Survival data of the glioma subgroup were extracted for analysis and generation of Kaplan-Meier curves for overall survival.

Results

Lower grade gliomas with less VWF gene expression had significantly better survival than those with more VWF gene expression (hazard ratio 0.64, 95% confidence interval 0.44-0.92, P 0.015 log rank test). The effect of VWF gene expression on survival was even more evident when the sample was analyzed as three groups (P = 0.00019). IDH1, TP53, and ATRX mutations are present in 40% or more adult LGGs.

Conclusion

The deleterious prognostic effect of VWF expression in LGGs is not surprising, given its role in other cancers. Therefore, VWF gene expression may be a clinically important risk marker in LGG.