Page 61 - GPD-3-4

Gene & Protein in Disease Drugs and immune infiltration in IPF

scale molecular networks and reveal gene interactions to identify and rank the hub genes. The hub gene network
and regulatory mechanisms. WGCNA analysis was was further visualized using Cytoscape_v3.10.1.
10
performed using the R language WGCNA package
(version 1.72-1). The top 25% ranked genes based on 2.6. Analysis and construction of the miRNA–TF–
variance in expression values were selected, outliers were mRNA network
removed, and a reliable WGCNA network was constructed. The miRTarBase, Starbase, and TargetScan databases
13
12
14
The soft-thresholding power was determined using the are widely used to predict miRNA–mRNA interactions,
“pickSoftThreshold” function. An adjacency matrix was enhancing our understanding of gene regulation. These
created and transformed into a topological overlap matrix databases contain extensive information on known
(TOM). Gene dissimilarity (1-TOM) was calculated, and miRNA–mRNA interactions, which can also be used to
the dynamic tree-cutting method was used to classify predict novel interactions. A Venn diagram was used to
genes into different modules. Modules with a dissimilarity identify overlapping miRNAs from all three databases.
coefficient of <0.2 (correlation coefficient of >0.8) were Enrichr (http://amp.pharm.mssm.edu/Enrichr/) is a web-
merged. The modules were further analyzed for correlation based platform for gene set enrichment analysis, offering
with clinical traits by calculating module membership a wide range of genomic libraries. The TRANSFAC and
(MM). This analysis revealed the module genes that JASPAR position weight matrix sections in Enrichr were
were most strongly associated with clinical traits, and the used to identify TFs regulating the expression of CGs using
correlation network between these key module genes and a P-value threshold of <0.05. After obtaining TF–mRNA
clinical traits was visualized. Finally, DEGs between IPF and miRNA–mRNA interaction data, they were integrated
and normal control samples were integrated with the key to establish the miRNA–TF–mRNA network. Cytoscape_
module genes to identify CGs as the final DEGs. v3.10.1 was then used to visualize this regulatory network.
2.4. Gene ontology (GO) and Kyoto encyclopedia of 2.7. Immune cell infiltration analysis
genes and genomes (KEGG) analyses of CGs for gene CIBERSORT is a gene expression analysis tool that uses
and protein annotation and pathway enrichment known gene expression data to identify and classify different

GO analysis is a bioinformatics method that uses ontology- cell types in a sample. It estimates the proportions of each
15
based approaches to categorize and annotate genes and cell type, providing insights into the distribution and gene
proteins into three main categories: cellular component expression characteristics of various cell populations. In this
(CC), biological process (BP), and molecular function study, we obtained expression profile data for 22 immune
(MF). KEGG is a comprehensive bioinformatics database cells from the CIBERSORT website (https://cibersort.
(https://www.genome.jp/kegg/) that provides extensive stanford.edu/). Using the CIBERSORT algorithm in R, we
information on genes and proteins, including gene quantified the relative proportions of infiltrating immune
sequences, protein structures, chemical reactions, and cells in IPF samples. We selected significant samples with
cellular signal transduction. Analyzing data using KEGG a P-value threshold of <0.05 and presented the results
helps in better understanding the functions of genes and using a bar plot. To visualize the distribution of immune
proteins as well as their roles in BPs. In this study, we used cell types in IPF samples, we constructed a heatmap using
the “clusterProfiler” R package (version 4.6.2) to perform the “pheatmap” package (version 1.0.12) in R. In addition,
GO analysis and KEGG pathway enrichment analysis of we used the corrplot package (version 0.92) to generate a
CGs. Statistically significant results were defined as an correlation heatmap illustrating the relationships between
adjusted P ≤ 0.05 and a minimum gene count of ≥2. infiltrating immune cells. Finally, we used the vioplot
package (version 0.4.0) to construct violin plots comparing
2.5. Analysis of PPI network and identification of the proportions of infiltrating immune cells between IPF
hub genes and normal control samples.
The STRING database is a specialized PPI database that
stores interactions from various species, including both 2.8. Screening of candidate drugs targeting hub
experimentally validated and predicted interactions. In genes
this study, the PPI network of CGs was analyzed using The CTD (https://ctdbase.org/) is a comprehensive
the STRING database with a minimum interaction score repository integrating information on chemical
cutoff of ≥0.4. Cytoscape_v3.10.1 was used to visualize substances, genes, functional phenotypes, and disease
the PPI network, and the CytoHubba plugin was utilized interactions, providing gene–drug–disease interaction
11
to identify hub nodes. Each gene in the PPI network was data to support drug screening. The DrugBank database
assigned a value using 12 topological network algorithms (https://go.drugbank.com/) integrates bioinformatics and

Volume 3 Issue 4 (2024) 3 doi: 10.36922/gpd.4101