Page 9 - TD-1-2
P. 9

Tumor Discovery                                                       BAK1 as a novel prognostic biomarker



            was performed on them. To assess the relationship between   Table 1. 33 cancer types used in this study
            high and low  BAK1 expression groups and immune cell
            infiltration content, the CIBERSORT algorithm was   Abbreviation  Full name
            used, and the filter condition for immune cell infiltration   ACC  Adrenocortical carcinoma
            findings was set to pFilter = 0.05. Immune checkpoint   BLCA  Bladder urothelial carcinoma
            correlation analysis was then performed to investigate the   BRCA  Breast invasive carcinoma
            relationship between immune checkpoint genes and BAK1,   CESC  Cervical squamous cell carcinoma and endocervical
            with the filter condition of correlation test P-value set to   adenocarcinoma
            pFilter = 0.001. The immunoscores of liver cancer patients   CHOL  Cholangiocarcinoma
            were gathered from http://tcia.at/, and immunotherapy   COAD  Colon adenocarcinoma
            analysis was performed in R studio program using “limma”   DLBC  Lymphoid neoplasm diffuse large B-cell
            and “ggpubr” packages. The immunohistochemical pictures   ESCA  Esophageal carcinoma
            of BAK1 protein in various cancer and normal tissues
            were obtained from http://www.proteinatlas.org/(Human   GBM   Glioblastoma multiforme
            Protein Atlas [HPA] database). We estimated the half   HNSC   Head and neck squamous cell carcinoma
            maximal inhibitory concentrations (IC50s) of compounds   KICH  Kidney chromophobe
            collected from the Genomics of Drug Sensitivity in Cancer   KIRC  Kidney renal clear cell carcinoma
            (GDSC) website in the TCGA project of the LIHC dataset   KIRP  Kidney renal papillary cell carcinoma
            to identify possible medications for clinical application in   LAML  Acute myeloid leukemia
            LIHC therapy. The IC50 of the chemical, acquired from   LGG   Low grade glioma
            the GDSC website, in LIHC patients was predicted using R   LIHC  Liver hepatocellular carcinoma
            studio’s “pRRophetic” package, with the filter condition for
            P-value set to pFilter = 0.001.                    LUAD       Lung adenocarcinoma
                                                               LUSC       Lung squamous cell carcinoma
            2.5. Building a nomogram scoring system            MESO       Mesothelioma
            The clinical variables and risk scores were integrated to   OV  Ovarian serous cystadenocarcinoma
            construct a predictive nomogram in R studio using “rms,”   PAAD  Pancreatic adenocarcinoma
            “survival,” and “regplot” packages. Each clinical feature and   PCPG  Pheochromocytoma and paraganglioma
            risk rating in the nomogram was assigned a score; the total   PRAD  Prostate adenocarcinoma
            score was calculated by summing the scores for all clinical   PEAD  Rectum adenocarcinoma
            features and risk ratings in each sample. The accuracy of
            anticipated 1-, 3-, and 5-year survival events was described   SARC  Sarcoma
            using calibration plots.                           SKCM       Skin cutaneous melanoma
                                                               STAD       Stomach adenocarcinoma
            2.6. Statistical analysis                          TGCT       Testicular germ cell tumors
            R version 4 was used for all statistical studies. Statistical   THCA  Thyroid carcinoma
            significance was defined as P < 0.05.              THYM       Thymoma
            3. Results                                         UCEC       Uterine corpus endometrial carcinoma
                                                               UCS        Uterine carcinosarcoma
            3.1. Expression and survival analysis of BAK1      UVM        Uveal melanoma
            Table 1 lists the 33 cancers studied in this work. The HCC
            cohort (GSE76427) used has 115  patients in total. We   in pan-cancer. Figure 1B shows that the expression of BAK1
            analyzed the PRGs in 115 HCC patients using R studio   was significantly upregulated in 11 cancer types (bladder
            packages “survival” and “limma.”  P < 0.05 was deemed   urothelial  carcinoma  [BLCA],  breast  invasive  carcinoma
            statistically significant for prognosis.  Figure  1A  depicts   [BRCA],  cholangiocarcinoma  [CHOL],  esophageal
            the gene coexpression relationship, revealing that GSDME,   carcinoma [ESCA], glioblastoma multiforme [GBM],
            CHMP4B, CHMP3, BAK1, and NOD2 are high-risk genes   head-and-neck  squamous  cell  carcinoma  [HNSC],  liver
            that are strongly associated with prognosis. In view of the   hepatocellular carcinoma [LIHC], lung adenocarcinoma
            limited number of research on the prognosis and immunity   [LUAD], lung squamous cell carcinoma [LUSC], stomach
            of  BAK1 expression in liver cancer, we chose  BAK1 and   adenocarcinoma [STAD], and uterine corpus endometrial
            used TIMER 2.0 database to investigate BAK1 expression   carcinoma [UCEC]; all  P < 0.001), whereas it was


            Volume 1 Issue 2 (2022)                         3                       https://doi.org/10.36922/td.v1i2.221
   4   5   6   7   8   9   10   11   12   13   14