Page 8 - TD-1-2

Tumor Discovery BAK1 as a novel prognostic biomarker

death with pro-inflammatory characteristics. It is classified 2.2. Gene expression and survival prognostic
into caspase-1-dependent classical pyroptotic pathway analysis
and caspase-4/5/11-dependent non-classical pyroptotic Using the “Diff Exp” module on the Tumor Immunity
pathway. Pyroptosis is characterized by deoxyribonucleic Estimation Resource (TIMER) website (http://timer.
acid (DNA) breakage, cell membrane rupture, and the cistrome.org/) and the R studio software, we investigated
release of pro-inflammatory proteins [5,6] . According BAK1 expression in 33 human tumor and normal control
to research, the expression of caspase-1 is low in liver tissues from the TCGA database. In LIHC, the packages
cancer tissue . Other research has demonstrated that “limma,” “ggplot2,” and “ggpubr” performed differential
[7]
hypoxia-induced caspase-1 activation and the subsequent and pairwise differential analyses on BAK1. Kaplan–
generation of different inflammatory factors in liver cancer Meier curves created by the R studio programs “survival”
tissues and cell lines can promote cancer cell invasion and “survminer” were used to analyze the differences in
and metastasis . Pyroptosis not only impedes tumor survival between subtypes. Univariate and multivariate
[8]
cell proliferation but also creates a microenvironment independent prognostic analyses were then carried out to
that promotes tumor cell development [9,10] . Given the determine if BAK1 could be used independently of other
importance of pyroptosis in malignancies, the aim of this prognostic indicators.
work was to identify HCC pyroptosis-related genes (PRGs)
and investigate their implications in HCC prognosis. 2.3. Clinical correlation analysis and coexpression
analysis
To identify prognosis-related pyroptosis genes, the
prognostic value of 52 PRGs in 115 HCC patients from the Clinical correlation analyses and heatmaps were created
Gene Expression Omnibus (GEO, GSE 76427) cohort was in R studio using “limma,” “ComplexHeatmap,” and
examined. Following the selection of BAK1 target gene, its “ggpubr” packages. Genes that share the same promoter as
expression level was obtained from The Cancer Genome BAK1 were identified. A correlation coefficient larger than
Atlas (TCGA) database. In addition to the construction of zero between the two indicates that the gene is positively
a nomogram, differential analysis, survival analysis, and regulated by BAK1, while a correlation coefficient
clinical correlation analysis were carried out to predict lesser than zero indicates that the gene has a negative
the survival rate. Subsequently, all samples were separated regulatory interaction with BAK1. The filter condition of
into two groups based on BAK1 gene expression: High the coefficient of correlation was corFilter = 0.6; the filter
and low expression. Enrichment analysis, immunological condition of the correlation test P-value was pFilter =
analysis, and drug sensitivity analysis were performed 0.001, and the coexpression circle graph was drawn based
on the differential genes. The role of BAK1 in predicting on the coexpression results.
prognosis and immunotherapy response in patients with 2.3. Gene enrichment analysis
liver cancer was investigated.
The samples were separated into two groups with high and
2. Materials and method low BAK1 expression levels, respectively, using “limma”
and “pheatmap” packages in R studio. A gene heat map with
2.1. Data sources
differences between the high and low expression groups
The clinically relevant data and gene expression of liver was generated. The logFCfilter parameter was set to 1, the
cancer were downloaded from TCGA database (https:// fdr filter condition was fdrFilter = 0.05, and the adjusted
portal.gdc.cancer.gov/). The GEO (https://www.ncbi.nlm. P-value was 0.05. We performed Kyoto Encyclopedia of
nih.gov/geo/) was also used in this work. For the following Genes and Genomes (KEGG) analysis of differential genes
analysis, a GEO HCC cohort (GSE 76427) and a TCGA in R studio program using “org.Hs.eg.db,” “clusterprofiler,”
cohort were collected. Thereafter, the transcriptome and and “enrichplot” packages to further investigate the
clinical data were combined and ID transformed. Fifty-two enrichment of probable pathways of differential genes in
pyroptosis-related genes (REACTOME PYROPTOSIS) different groups. To further explore the enrichment of
were obtained from previously published studies and potential pathways of differential genes in different groups,
the Molecular Signatures Database (MSigDB) (http:// we performed KEGG and GSEA enrichment analysis.
www.broad.mit.edu/gsea/msigdb/) [11,12] . TCGA and GEO
data were integrated in R studio using “limma” and “sva” 2.4. Immune correlation analysis and drug
packages, and the expression of the PRGs was retrieved sensitivity analysis
from the merged data. Finally, the survival analysis of Through differential analysis of immune cells, immune cells
pyroptotic genes was performed to obtain the prognosis- with statistical significance between high and low BAK1
related pyroptotic genes. expression groups were discovered, and differential analysis

Volume 1 Issue 2 (2022) 2 https://doi.org/10.36922/td.v1i2.221