Analysis of Gene Expression Omnibus high-throughput sequencing data for the determination of microribonucleic acids in the blood plasma of patients with glioblastomas

Abstract

Purpose of the study. This work is devoted to the study of blood plasma miRNA patterns in blood plasma using high-throughput sequencing of the Omnibus Gene Expression base and the search for candidate miRNA molecules for the development of a minimally invasive diagnostic panel.Materials and methods. Basing on the open dataset of Omnibus Expression of the NCBI GSE150956 Gene, groups of samples with glioblastoma and conventionally healthy donors were formed. For each sample, information on the levels of miRNA expression was extracted. Determination of significant miRNAs using machine learning algorithms of the R 4.0.4 project. For significant miRNAs, target genes have been performed, an analysis of the improvement of functional characteristics and interactome analysis of target genes of miRNA were performed.Results. The study analyzed the data of 131 samples, where 35 samples with glioblastoma and 96 samples of the conditionally healthy group. Differential expression data were obtained for 945 miRNA. Two panels were obtained using machine learning methods, common miRNA – hsa-miR 3180, hsa-miR 3180-3p, hsa-miR 6782-5p, hsa-miR 182-5p, hsa-miR 133b and hsa-miR 670-3p. For significant miRNAs, information was obtained on experimentally confirmed target genes, a gene ontology demonstrating their participation in enzyme binding, participation in the regulation of primary cellular metabolic processes, and the development of glioblastomas and cancer in general.Conclusion. As a result of layer-by-layer filtering and application of machine learning algorithms, significant miRNAs were identified that are candidates for a diagnostic panel of a minimally invasive method of high-grade glial tumors.

Keywords

• glioblastoma
• circulating microrna
• blood plasma
• next-generation sequencing
• network analysis
• biomarker