Page 41 - Read Online

Aydin et al. J Transl Genet Genom. 2025;9:406-26 https://dx.doi.org/10.20517/jtgg.2025.108 Page 410

Molecular docking analysis to determine efficiency of repositioned drugs in silico
Protein Data Bank (PDB) and UniProt were used to select the structures of hub proteins. Protein
[29]
[30]
structures were selected based on X-ray crystallography data with a resolution of ≤ 1.8 Å, as higher-
resolution structures provide more accurate atomic positioning and improve the reliability of subsequent
docking analyses. Inhibitors of each diagnostic biomarker were searched through the Comparative
[31]
Toxicogenomics Database (CTD) from the chemical-gene interaction section. The inhibitors were
selected based on their ability to decrease the expression of the diagnostic biomarker genes. Three-
dimensional structures of inhibitors and drugs were downloaded using PubChem . The clinically used
[27]
DMD medication vamolorone (VBP15) was also included in the screening pipeline to benchmark our
candidate compounds. All ligands, including vamorolore, were subjected to the same preparation workflow
and docked against the hub proteins using the CB-DOCK (cavity-detecting blind docking) blind docking
platform . This allowed direct comparison of binding affinities between FDA-approved therapeutic agents
[32]
and the repositioned drug candidates identified in this study. Molecular docking analysis was performed
using CB-Dock .
[32]
RESULTS
Identification of DEGs for DMD
The datasets with accession numbers GSE70955, GSE38417, and GSE109178 were chosen for understanding
the molecular mechanism of DMD at the transcriptome level. With the aid of GEO2R, the test and control
groups for every set were established, and their differential expression was determined. The significance
thresholds were set at a P-value < 0.001 and an |log fold change| ≥ 1.0. According to the differential
2
expression analysis, 971 DEGs were identified in the GSE70955 dataset, 3,915 in the GSE38417 dataset, and
6,410 in the GSE109178 dataset. A comparison of the number of DEGs across the selected datasets shows
that GSE109178 yielded the highest count. This dataset has a relatively large sample size, which increases
statistical power, provides technical robustness, and leads to more genes reaching significance under a
stringent threshold (P < 0.001, |log FC| ≥ 1). The variation in the number of DEGs across the selected
2
datasets may arise from differences in patient cohort characteristics - such as disease severity, age, muscle
type biopsied, and treatment history - which collectively broaden or constrain overall transcriptional
variability. Among the DEGs from the three datasets, overlapping genes were identified, resulting in 285
genes shared across all datasets [Figure 1A].

Functional enrichment analysis resulted in biological pathways that play crucial roles in
pathogenesis
The common DEGs derived from specific datasets (GSE38418, GSE70955, and GSE109178) were
functionally enriched and found to be significantly associated with extracellular matrix organization, muscle
structure development, and cytoskeleton in muscle cells, as demonstrated by the Metascape results
[Figure 1B]. Dark colors were used to express the first two processes, whereas less intense colors were used
for the others. Darker hues often indicate more expressed genes or proteins, while lighter hues indicate less
expressed ones .
[17]

Networks constructions revealed DMD-specific molecular signatures
Network construction for DMD was performed using three different approaches: PPI, TF, and miRNA
analyses. A total of 285 common DEGs identified from the GSE38417, GSE70955, and GSE109178 datasets
were used. The PPI network comprised 1,739 nodes, 3,961 edges, and an average of 4,555 neighbors; the TF
network included 94 nodes, 122 edges, and 2,596 average neighbors; and the miRNA network contained
2,189 nodes, 8,752 edges, and 6,095 average neighbors.