Page 280 - EJMO-9-2

Eurasian Journal of
Medicine and Oncology QGJSF multi-target mechanisms in osteoporosis

druggable score ≥0.1. Targets obtained from both databases 2.7. Molecular docking
were combined, and duplicate entries were removed to Molecular docking was used to evaluate the binding affinity
ultimately acquire the potential targets of the QGJSF. between active ingredients and their target proteins. The

2.2. Acquisition of OP Targets 3D structures of key active components from the QGJSF
were downloaded from the PubChem database, and the 3D
Disease targets related to OP were retrieved by searching structures of core target proteins were obtained from the
the GeneCards (https://www.genecards.org/), OMIM Protein Data Bank (PDB) (https://www.rcsb.org/). Protein
(https://omim.org/), and DisGeNET (https://disgenet. receptors were pre-processed using PyMOL software to
com/) databases using “osteoporosis” as the search term. remove water molecules and ligands. Subsequently, the
2.3. Construction of PPI network for key targets of key components and target proteins were imported into
QGJSF in treating OP AutoDockTools 1.5.7 for hydrogen atom addition and
active site determination. The docking binding energy was
Targets related to the treatment of OP by QGJSF were then calculated. A binding energy of <0 kcal/mol indicates
uploaded to the online platform Wei Sheng Xin website that the receptor and ligand can bind spontaneously, while
(https://www.bioinformatics.com.cn/). Using the a binding energy <−5 kcal/mol suggests favorable binding
interactive Venn diagram tool, the intersection between activity.
QGJSF-associated targets and OP-related molecules was
identified, which represents the potential targets of QGJSF 2.8. GEO database validation
for OP. These potential targets were then submitted to the The “limma” R package was used to analyze the differential
STRING database (Version: 11.0, https://string-db.org/) gene expression in the OP dataset GSE5958 from the GEO
for PPI analysis, with the species limited to Homo sapiens database (https://www.ncbi.nlm.nih.gov/geo/). Genes
and the confidence score threshold set to the highest level meeting the criteria of |logFC| >1 and an adjusted p<0.05
(0.900), excluding isolated nodes. were considered significance differentially expressed.
2.4. Screening for core targets Volcano plots and heatmaps were generated using the
“ggplot2” and “pheatmap” R packages, respectively. A Venn
The mapped targets obtained from the screening process diagram was utilized to identify common targets between
were imported into the STRING database (https:// the potential targets of Strychni Semen and the GSE5958
cn.string-db.org/) to acquire the TSV file of the PPI dataset. The expression of core genes was visualized and
network. This file was then imported into Cytoscape 3.9.0 validated using dataset GSE35958 with the “ggpubr”
software, and network analysis was performed using the package. Subsequently, receiver operating characteristic
Centiscape 2.2 plugin. The significance of each potential (ROC) curves were constructed using the “pROC” package
target was evaluated across three centrality measures: to evaluate the predictive performance of marker genes.
Degree, betweenness, and closeness. The intersection of
the top-ranking targets was identified as the core targets, 2.9. GSEA
which were subsequently visualized for further analysis. In this study, single-gene GSEA analysis was conducted

2.5. Construction of the “Drug-Target-Disease” network using the “gseaplot2” package to investigate the potential
functional role of the identified hub gene.
The core targets associated with the treatment of OP by
QGJSF were organized and imported into Cytoscape 3.9.0 3. Results
to construct a “TCM-Active Component-Target” network,
which was subsequently visualized for further analysis. 3.1. Screening of targets related to QGJSF
The targets related to the QGJSF were retrieved from
2.6. GO Function and KEGG Pathway Enrichment TCMSP and BATMAN-TCM databases. The identified
Analysis of Core Targets targets were converted using the UniProt database, and
GO functional annotation and KEGG pathway enrichment duplicate entries were removed, resulting in a total of 1,395
analysis were performed on the core targets through the potential targets for the QGJSF.
DAVID database (Version: 6.8, https://david.ncifcrf.gov/),
with both species and background set to Homo sapiens. 3.2. Acquisition of OP-related targets and common
The top 20 primary BP and non-disease, non-cancer- targets between drug and disease
related signaling pathways with pharmacological relevance A total of 2,784 OP-related targets were obtained from
(p<0.05) were selected to explore the potential mechanisms the GeneCards, OMIM, and DisGeNET databases. By
by which QGJSF may treat OP. intersecting the predicted targets of the QGJSF with

Volume 9 Issue 2 (2025) 272 doi: 10.36922/EJMO025150103