ABSTRACT
Background: Gliomas are the most common primary intracranial tumor, containing 81% of brain malignant carcinoma. Glioblastoma is a primary brain tumor with a very poor prognosis. Existing treatment methods have many limitations, so more in-depth researches are urgently needed. Results: : To screen key genes and build models, we made full use of RNA sequencing (RNA-seq) data from TCGA and GTEx and perform immune infiltration analysis in multiple immune-related databases. Finally, we screened key genes and build the model. Meanwhile, we explored immune cell infiltration. KEGG and GO analysis prompted us COVID-2019 may have potential relation with GBM. Further selection of immune-related differential genes and gene-pairs were then explored. The prognosis model built by lasso regression was tested to have great effect. Key genes were all found to have quite diverse expression level compared with normal samples. Moreover, their co-expression analysis reminded us more attention could be paid to WNT and NOTCH signaling. Immune infiltration suggests that most immune cells are differentially expressed in tumor samples, and the interaction reminded us B cells may play a critical role. Conclusion: This study is the first to build immune-gene-pairs and used lasso regression to construct the model and screen critical pairs. Key genes were then selected and observed their role in GBM.