Skip to main content
NCBI home page
PLOS One logoLink to PLOS One
. 2026 Feb 20;21(2):e0343274. doi: 10.1371/journal.pone.0343274

System biology and network-based approach to identify the therapeutic signatures and potential inhibitors against polycystic lipomembranous osteodysplasia with Sclerosing Leukoencephalopathy

Bayan T Bokhari 1,*, Alaa M Saleh 1, Hashim M Aljohani 2,3,4, Wejdan Hussain Owaydhah 5, Mohammad Ahmad Alobaidy 6, Naief Dahran 7, Hind M Naffadi 8, Alaa Abdulaziz Eisa 2,4,*
Editor: Md Zubbair Malik9
PMCID: PMC12923142  PMID: 41719345

Abstract

Polycystic Lipomembranous Osteodysplasia with Sclerosing Leukoencephalopathy (PLOSL), often recognized as Nasu-Hakola disease, is an uncommon autosomal recessive systemic condition described by numerous bone lesions that resemble cysts and progressive early-onset dementia. One novel way to find possible drug targets is to use Network Analyst tools for network-based gene expression profiling. Identifying the target hub genes, essential for the initiation and progression of PLOSL, is validated by the significance level (p-score) attained using Cytoscape in the survival analysis of the major central genes. The X2K online tool also examined the regulatory kinases that formed the interaction between protein molecule networks. Out of the 53 genes obtained to be differentially expressed, PPARG and AIF1 had the greatest degree score, followed by C1QA with 5 degrees and SIGLEC1 and MSR1 with 4 degrees. Furthermore, Molecular docking of target PPARG gene with AMG-131 and Elafibranor drugs, having the chemical formulas 2-[2,6-dimethyl-4-[(E)-3-(4-methylsulfanylphenyl)-3-oxoprop-1-enyl]phenoxy], and 3,5-dichloro-4-quinolin-3-yloxyphenyl) benzenesulfonamide, along control (Rosiglitazone (S) shows binding affinities of −7.2 kcal/mol, −7.4 kcal/mol, and −7.6 kcal/mol respectively. The enzyme remained extremely stable in the complex throughout 200 ns, with a mean Root Mean Square Deviation (RMSD) of 2.95 Å for the AMG-131 complex system and 2.93 Å against the Elafibranor complex system. Root Mean Square Fluctuation (RMSF) anticipated steady behavior with average RMSD for the active site residues in the docked system. Arg76 and Leu28leu Leu118 were shown to be essential enzyme residues for binding, anchoring, and bridging strong hydrogen and hydrophobic interactions between the enzyme and the inhibitor, according to the Radial Distribution Function (RDF). These results broadened our knowledge of putative biomarkers for PLOSL diseases, and an experimental strategy will improve our results even more in the future.

1. Introduction

Nasu-Hakola disease, also known as polycystic lipomembranous osteodysplasia with sclerosing leukoencephalopathy (PLOSL), is an uncommon autosomal recessive condition marked by a combination of irreversible dementia with early onset and polycystic bone degradation [1]. Due to its strong genetic and hereditary base, the disease is more common in some populations, such as Finnish and Japanese groups [2]. Proper phagocytosis and osteoclast proliferation are disrupted by mutations in TREM2 or TYROBP, which are crucial regulators of immunological signaling and microglial function. This results in severe neurological and muscular signs characteristic with PLOSL, as well as neuroinflammation and skeletal fragility [1,3,4]. It is crucial to comprehend the disease’s molecular underpinnings and find therapeutic targets because it progresses quickly and the only treatments available are supportive and symptomatic care [5].

Drug research is still expensive and time-consuming; typical pipelines need a significant financial investment and more than ten years [6,7]. While early-stage discovery capacity has increased because to advancements in combinatorial chemistry and high-throughput screening methods, authorized therapies have not increased proportionately [8]. This emphasizes the necessity for alternate tactics that can rank potential targets in order of importance and weed out unsuitable applicants early on [9]. Computational methods, such as structure-based drug design (SBDD) and computer-aided drug design (CADD), provide effective tools for assessing molecular interactions, forecasting pharmacological response, and directing early discovery efforts [10,11]. Simultaneously, systems biology integrates transcriptome, proteomic, and network-level data to provide a strong foundation for comprehending complicated disorders.

In this work, we characterize the molecular landscape of PLOSL using a combined systems biology and computational modeling approach. We want to uncover important genes, pathways, and possible medicinal drugs using molecular docking, molecular dynamics simulations, protein–protein interaction (PPI) network modeling, and differential expression analysis. By generating novel hypotheses and emphasizing prospects for future experimental validation, this integrated strategy aims to address the existing paucity of tailored treatments for this uncommon and crippling condition.

2. Materials and methods

This study incorporated an integrative approach combining systems biology and network analysis, as outlined in Fig 1, to elucidate molecular pathways and potential therapeutic targets for PLOSL.

Fig 1. Systematic representation of the integrated bioinformatics analysis utilized in this study.

Fig 1

Open in a new tab

2.2. Data acquisition for comparative analysis

We sourced our GSE3624 dataset from the NCBI’s GEO database (https://www.ncbi.nlm.nih.gov/geo/), a repository for high-throughput genomic data [12]. This dataset comprised eight samples, three from normal individuals and five from PLOSL patients, all analyzed on the [HG-U133A] Affymetrix Human Genome U133A Array platform [13].

2.3. Delineation and screening of Differentially Expressed Genes (DEGs)

The Network Analyst platform (https://www.networkanalyst.ca/) evaluated gene expression variations [14]. Comparing the expression patterns of bone marrow cells from patients with PLOSL with normal bone marrow cells proved to be a helpful application of this technology. The selection of genes was based on statistically significant differences. The data matrix was set up so that separate rows represented different genes and corresponding columns represented samples. Three control samples and five PLOSL patient samples were included in this matrix. To ensure consistency and facilitate analysis, gene identifiers were methodically transformed into Entrez IDs [15].

A Network Analyst for a meta-analysis of microarray data was used, applying Variance Stabilizing Normalization (VSN) using maximum likelihood estimates and parametric modifications to align sample variances regardless of their mean intensities. Log2 and quantile normalization were also used, with box plots and PCA plots verifying their effectiveness [16]. Differentially expressed genes (DEGs) were identified using strict criteria: adjusted p-value (FDR) < 0.05 (Benjamin-Hochberg method), a t-test based on Pearson correlation and the Limma algorithm to investigate the relationships between various variables [17]. Compared to normal controls, this multimodal approach yielded a comprehensive list of DEGs in PLOSL samples.

2.4. Integration of gene ontology and pathway enrichment

For the gene set functional enrichment study, Gene ontology (GO) components and KEGG pathway linkages were the main areas of interest which were conducted using TNMplot and Network Analyst [14]. These investigations made the molecular and cellular processes impacted by PLOSL clearer. We have used distinct subgroups of GSE3624, which had eight samples in total: Three from normal individuals and five from PLOSL patients. To find differentially expressed genes (DEGs) that were pertinent to our goal, the full dataset was processed using the Network Analyst tool. To evaluate the KEGG pathway and Gene Ontology (GO) enrichment of differentially expressed genes (DEGs) with a significant level of p < 0.05, the free bioinformatics tools TNMplot (https://tnmplot.com/) and NetworkAnalyst (https://www.networkanalyst.ca/) were utilized [18]. DEG modules and central genes were uploaded for functional and pathway enrichment analysis, considering a number of parameters, including cell components, molecular functions, biological processes, function mode, gene ID, species selection (Homo sapiens), and supporting information from pertinent GO and KEGG databases. Large-scale genomic or transcriptional datasets can be interpreted for their functional roles using the well-known Gene Ontology (GO) analysis approach [19].

2.5. PPI network generation

The STRING database (https://string-db.org/) [20] was targeted to map the interactions of proteins linked to the identified DEGs in PLOSL, which contains data on over 67 million proteins across approximately 14,000 species. Cytoscape 3.9.1 visualized these molecular networks and biological pathways [21], allowing us to import the interaction data from STRING. A confidence threshold of 0.75 was set, ensuring medium confidence in the network’s validity [18,19], focusing on the original DEGs. The entire dataset, encompassing these genes, was integrated into Cytoscape [22]. To avoid complexity, zero-order interaction was used to prevent overwhelming complexity, commonly known as the ‘hairball’ effect [23]. The “Network Analyzer” feature in Cytoscape further assessed parameters like network density, node degree distribution, clustering coefficient, power law distribution of nodes, and network centralization [24], providing insights into key protein interactions relevant to PLOSL’s pathophysiology.

2.6. Key hub proteins identification in the PPI network

The Cytoscape (https://cytoscape.org/), a sophisticated tool, was used to visualize molecular interaction networks and identify key hub proteins in the PPI network [22]. Protein centrality was calculated based on the number of connections or edges that a protein shared with other proteins in the built PPI network. These hub genes are important and potentially crucial to PLOSL’s molecular mechanisms. Hub gene candidates were identified as proteins with a high degree of connectedness, which indicated a significant number of interactions within the network [25]. Incorporating eleven distinct algorithms, the CytoHubba plugin in Cytoscape helped identify critical nodes and narrow the focus to the most significant genes in our PPI network [26]. Due to their significant connectivity, these central hub proteins are hypothesized to be vital in the pathophysiology of PLOSL and, therefore, might emerge as promising targets for therapeutic interventions.

2.7. Analysis of regulatory associations and key transcription factors

Key transcription factors (TFs) and kinases were identified using the Expression2Kinases tool (https://maayanlab.cloud/X2K/), influencing the identified DEGs. The entire array of DEGs, complete with specified gene markers, was uploaded to the X2K system for analysis [27]. This tool adeptly calculated the ten most prominent transcription factor regulators and associated kinases, utilizing its kinase module and transcription factor (TF) analysis. The assessment was grounded on the Fischer test p-values, ensuring a robust statistical framework for the findings. X2K facilitated the construction of a comprehensive regulatory network by integrating data from the ChEA69 database, which is rich in ChIP-seq-derived information. The resulting network was shown using Cytoscape’s capability to read and display “graphml” files. This dynamic model visualization captured the intricate interactions between different transcription factors and kinases, not just a static depiction [27].

2.8. Survival analysis of hub genes

The Kaplan-Meier plotter program (https://kmplot.com/), a versatile online tool [28] generates survival curves for each hub gene, demonstrating the relationship between the expression of genes and survival rates of patients. This approach provides a mathematical and visual representation of survival probabilities over time, aiding in identifying potential prognostic biomarkers for PLOSL.

2.9. Protein-drug interaction network

The possible medication interactions with top hub genes using Drug Bank (version 5.1.10) was used, an online database integrated with Network Analyzer. This analysis included FDA-approved drugs and investigational compounds. This extensive database encompasses many substances ranging from FDA-approved drugs to compounds that are still in the investigational phase. The Drug Bank database is remarkably extensive, featuring a total of 15,466 drug entries. This includes 2,741 small molecule drugs that have received FDA approval, 1,577 biologics such as proteins, peptides, vaccines, and allergens, alongside 134 nutraceuticals, and a significant number of drugs (over 6,717) currently in the experimental stage. Furthermore, the database links these drug entries to 5,293 distinct non-redundant protein sequences, which serve various roles, such as drug targets, enzymes, transporters, or carriers [29].

2.10. Data acquisition for comparative analysis

The NCBI Gene Expression Omnibus (GEO) database was sourced from a publicly accessible gene expression data repository (https://www.ncbi.nlm.nih.gov/geo/), The datasets used in this work, GSE3624 and GSE25496. Eight samples in all three from normal individuals and five from PLOSL patients are included in the GSE3624 collection. while the GSE25496 dataset, on the other hand, includes two samples, one from a PLOSL patient and one from a healthy control. The Affymetrix Human Genome U133A Array (HG-U133A), a microarray platform frequently used for gene expression profiling, was utilized to create both datasets. Both datasets were selected as the most pertinent and easily accessible sources for this uncommon illness since they are the only publicly accessible transcriptome datasets with PLOSL patient samples and matched controls.

2.11. Molecular docking of PPARG gene with target drugs

To find the most effective inhibitors, molecular docking analysis was performed using PyRx software [30]. We have retrieved the target receptor PPARG involved in Lipomembranous Osteodysplasia with Sclerosing Leukoencephalopathy (PLOSL), from the protein database with PDB id: 1FM6. Receptor active sites were predicted using literature support [31]. An Arg76 hotspot residue that is thought to be critically important for catalytic activity was used as the coordinates for docking against the top 2 compounds, i.e., AMG-131 and Elafibranor. The dimensions of the binding site at 15 Å, which had an X-axis of 16.777, Y-axis of −0.086, and Z-axis of 3.233, were determined by pointing coordinates of the atom Arg76. This has been followed by positive and negative control inhibitors with other FDA approved drugs, i.e., Positive Control- Rosiglitazone (S), Samarium Sm-153 Lexidronam, Clemastine, and Negative control Acetaminophen against their respected targets. Eight iterations were allocated to each compound, and the compound that achieved the largest negative binding energy (kcal/mol) and the most stable binding conformer was ranked first. MD simulations were performed for both complex systems [33].

2.12. Molecular Dynamics (MD) simulation

The molecular docking prediction regarding the lead compounds achieved during the network pharmacological study with its binding mode was assessed using MD simulation. The trajectories generated during the MD simulation production phase, the stability data and the profile of chemical interactions of the MD system were recovered. MD simulations were run using the AMBER 16 simulation program [32]. The Antechamber software produced the MD system parameters library [33]. The compounds’ parameters were created using the General Amber Force Field (GAFF) [34] whereas, FF14SB force field was used for protein [35], while the AMBER leap module [36] helped center the complex in a 12 × TIP3P water box [37]. The system’s charge was neutralized by adding the proper counter ions. The energy of the system was gradually reduced through multiple rounds: 100 steps were dedicated to minimizing hydrogen atoms, 500 steps to reducing water boxes, 500 steps to minimizing the entire system while imposing a 5 kcal/mol Å2 constraint on Cα, and 100 steps were devoted to minimizing non-heavy atoms while maintaining a 10 kcal/mol Å2 constraint. Next, an NVT ensemble was used to gradually heat the system from 0 to 300 K over 20 ps, with a time step of 2 femtoseconds and a constraint of 5 kcal/mol –Å2 on systems Cα. While the SHAKE algorithm [38], a constant hydrogen bond length and langevin dynamics [40] were maintained to accomplish temperature hold. Then, using a time step of two femtoseconds, the systems were adjusted for 1 ns. After that, each system’s constant temperature and pressure (NPT) ensemble [39] was created by running the system for 1 ns at 300 K and 1 bar of pressure. At last, a 1 ns equilibration period was granted to the systems. For the lead compound, a 200-ns production run was carried out. Using a time step of 2 fs and non-bound interactions set to 8 Å, herein the production phase was initiated within NVT ensemble, where a Berendsen temperature coupling method was used to maintain a thermal stability constant temperature [40]. The CPPTRA module was utilized to conduct simulation trajectories to investigate structural systems [41].

3. Results

3.1. Quality control and principal component analysis

Expression profiling was performed for another data set for comparative analysis using the PLOSL gene datasets GSE3624 and GSE25496 from GEO (NCBI), with each subject detail presented in Table 1 and Table 2. To identify DEGs, p-values and log2 fold change (log2FC) values were used by the Network Analyst tool, considering genes with logFC > 1 or <−1 and p-value < 0.05 as significant, with DEGs calculated using the predetermined cutoff criteria. Quantile normalization like variance-stabilizing normalization, was applied to assess transcriptome expression, and normalization results are shown in S1 Fig. Principal component analysis (PCA) identified unique groupings based on gene transcription levels in mutant and control samples.

Table 1. Displays the data related to the samples in the GSE3624 dataset that were retrieved from the Gene Expression Omnibus database.

Group Accession Organism Disease State Cell type
Normal GSM626950 Homo sapiens Normal The control brain RNA
PLOSL GSM626951 Homo sapiens PLOSL The Nasu-Hakola disease brain RNA
Open in a new tab

Table 2. Presents data pertaining to the samples in the GSE25496 dataset obtained from the Gene Expression Omnibus database.

Group Accession Organism Disease State Cell type
Normal GSM83892 Homo sapiens Normal Monocyte-derived DCs
GSM83894 Homo sapiens Normal Monocyte-derived DCs
GSM83894 Homo sapiens Normal Monocyte-derived DCs
PLOSL GSM83889 Homo sapiens PLOSL Monocyte-derived DCs
GSM83890 Homo sapiens PLOSL Monocyte-derived DCs
GSM83891 Homo sapiens PLOSL Monocyte-derived DCs
GSM83895 Homo sapiens PLOSL Monocyte-derived DCs
GSM83896 Homo sapiens PLOSL Monocyte-derived DCs
Open in a new tab

3.2. Differential expression analysis and PPI network analysis

A thorough study was conducted on the normal and unhealthy groups of both the data set GSE3624 and GSE25496 using several statistical tests, such as the student’s t-test, the Pearson correlation test and the Benjamin-Hochberg technique. The investigations conducted yielded a list of 53 DEGs in GSE3624 data set while 20 DEGs in GSE25496. Among these discovered DEGs, 41 genes showed overexpression, and 12 genes showed downregulation in GSE3624 data set, on other hand 8 genes show upregulation and 12 genes show downregulation in GSE25496 data set were finally prioritized based on the changed or corrected p-value. A key strategy for revealing the complexities of cellular networking systems is to investigate the PPI network. The degree of the ailment can be inferred from differences seen in the protein networks of cells between the normal and unhealthy groups. Nodes in this computer model represent genes and proteins, while edges show the links or connections between these genes and proteins, by mapping the discovered DEGs, a network representing the relationships between them was created. Cytoscape was used to display the visualization of these discovered DEGs. A moderate confidence level of 0.75 was used for both the data sets to query the String repository to obtain interactions between DEGs. For GSE3624 dataset there were 28 nodes and 27 edges in the network. Calculations revealed that the mean local clustering coefficient was 0.377, and the mean node degree was 1.02. On the other hand, For GSE25496 dataset there were 12 nodes and 20 edges in the network, local clustering coefficient was 0.186 and the mean node degree was 1. Furthermore, Genes were represented as nodes in this study, with edges representing their connections. The differentially expressed genes (DEGs) discovered in the investigation were used to design an interaction network. These DEGs were mapped, and their associations were displayed with Cytoscape. Data for these associations were acquired from the STRING database, with a moderate confidence level of 0.75. These measures show the significance of the relationships inside our network in Fig 2 and S2 in S1 Fig. In addition, Fig 3 and S3 in S1 Fig depicts the DEGs visually using a volcano plot and a heat map. The volcano plot demonstrates the statistically significant and fold-change pattern of the DEGs, whilst the heatmap depicts expression patterns across samples, allowing for clear separation between upregulated and downregulated genes. These results, taken together, provide evidence for the dependability and biological importance of the detected DEGs.

Fig 2. There were 27 edges and 28 nodes in the network that was built.

Fig 2

Open in a new tab

The analysis’s topological measures revealed robust interconnections within the network, with a mean node value of 1.02 and a mean local cluster coefficient of 0.377. The PPI data of the differentially expressed genes (DEGs) were visualized using Cytoscape software and the STRING database.

Fig 3. Displaying the intensity distribution of each gene on a volcano map, with significant downregulated, upregulated genes in the specimens colored in grey based on an adjusted p-value (FDR) < 0.05, and upregulated genes indicated in red and downregulated genes in blue.

Fig 3

Open in a new tab

(B) Using a heat map to display genes according to their expression levels across all samples. The distribution of expression is shown by the legend, which goes from low (blue) to high (red).

3.3. GO function and KEGG pathway enrichment function analysis

To determine the top 10 genes from the enrichment study of both the upregulated and downregulated top 10 target hub genes, GO analysis was carried out using the TNMplot and Network Analyst tools. These GO findings encompassed categories such as biological processes, molecular activities, cellular components, and KEGG pathways. One particularly relevant criterion used in the analysis was the false discovery rate (FDR). The top 10 target hub genes associated with the cellular component (CC) category included (GO:0005581-collagen trimer), (GO:0090575-RNA polymerase II complex), (GO:0048471-perinuclear region), (GO:0005938-cell cortex), and (GO:0043235-receptor complex). For molecular function (MF), two important protein roles were identified: (GO:0046790-virion binding) and (GO:0001540-beta-amyloid binding). Moreover, the identification of augmented biological processes (BP) in the investigation of DEGs includes (GO:0010629-negative regulation of gene expression), (GO:0042953-lipoprotein transport), (GO:0001774-microglial activation), (GO:0006954-inflammation), (GO:0045429-nitric oxide biosynthesis), (GO:0045087-innate immune response), (GO:0006911-phagocytosis), and (GO:0001666-response to hypoxia). Finally, the KEGG pathways include (hsa04380-Osteoclast differentiation) and (hsa04148-Efferocytosis). Functional details and pathways are listed in Table 3.

Table 3. Shows the overview of DEGs distribution in highly significant cellular components, molecular functions, biological processes and KEGG pathways based on FDR significant value.

GO_ID Term description Rate of False discovery Matching proteins in the network (labels)
Cellular components
GO:0005581 collagen trimer 0.040243823 MSR1, C1QA
GO:0048471 perinuclear region of cytoplasm 0.040433852 PSEN2, PPARG, AIF1
GO:0090575 RNA polymerase II transcription factor complex 0.05278098 PPARG, ATF3
GO:0005938 cell cortex 0.075794764 PSEN2, ITPR2
GO:0043235 receptor complex 0.09392639 ITPR2, PPARG
Molecular function
GO:0046790 virion binding 0.005680907 PTX3, SIGLEC1
GO:0001540 beta-amyloid binding 0.04054124 MSR1, C1QA
Biological processes
GO:0010629 negative regulation of gene expression 0.0089432 MSR1, PPARG, AIF1
GO:0042953 lipoprotein transport 0.009216997 MSR1, PPARG
GO:0001774 microglial cell activation 0.013797115 C1QA, AIF1
GO:0048662 negative regulation of smooth muscle cell proliferation 0.015167476 PPARG, AIF1
GO:0006954 inflammatory response 0.016256342 PTX3, SIGLEC1, AIF1
GO:0045429 positive regulation of nitric oxide biosynthetic process 0.021540137 PTX3, AIF1
GO:0045087 innate immune response 0.03279628 C1QA, PPARG, PTX3
GO:0006911 phagocytosis, engulfment 0.052415886 MSR1, AIF1
GO:0001666 response to hypoxia 0.078557149 PSEN2, ITPR2
KEGG Pathways
hsa04380 Osteoclast differentiation 0.075704469 ITPR2, PPARG
hsa04148 Efferocytosis 0.087056657 C1QA, PPARG
Open in a new tab

3.4. Hub genes detection

For both the target data set GSE3624 and GSE25496 the relationship between the hub genes and the nodes was assessed and categorized. Red-colored nodes show strong connections. Nodes with a degree value greater than 6 were considered hub nodes using Cytohubba. Additionally, using CytoHubba, which also computes the eccentricity, betweenness Centrality, and closeness centrality. (see Table 4 and S1 Table in S1 Fig). Notably, with a score of 6, for GSE3624 dataset PPARG showed the highest degree while for GSE25496 dataset AIF1 showed the highest degree score. The hub gene interactions’ shortest path network such as, degree score, eccentricity, betweenness Centrality, and closeness centrality are indicated by colors in Fig 4 and Fig S4 in S1 Fig. The rare neurological disease Polycystic Lipomembranous Osteodysplasia with Sclerosing Leukoencephalopathy (PLOSL), which is associated with lipid metabolism abnormalities, is largely caused by the PPARG gene. Elafibranor and AMG-131) are promising PPARG-targeting medications that aim to restore metabolic balance.

Table 4. The STRING protein-protein interaction database determines the top 10 genes found through the DEGs network analysis. The fold changes and corrected values for each gene are also included in the table.

Rank Gene name Score logFC Adjusted P-value Eccentricity Betweenness Centrality Closeness Centrality
1 PPARG 6 −1.8855 0.00015343 5 0.013888889 0.391304348
2 AIF1 6 3.1108 0.00000034596 3 0.333333333 0.642857143
3 C1QA 5 1.8134 0.000027186 5 0 0.3
4 SIGLEC1 4 2.4926 0.00011945 4 0 0.409090909
5 MSR1 4 −1.3591 0.00023267 4 0.222222222 0.409090909
6 MS4A6A 3 2.3256 0.0000044162 4 0.152777778 0.529411765
7 PSEN2 2 −0.72774 0.00022877 3 0.319444444 0.642857143
8 ITPR2 2 0.99659 0.0000067195 4 0 0.5
9 PTX3 2 −1.231 0.00016209 3 0.388888889 0.5625
10 ATF3 2 −1.6259 0.0000090115 4 0.041666667 0.45
Open in a new tab

Fig 4. All hub genes found have their shortest path interactions shown here, with nodes colored according to their degree scores.

Fig 4

Open in a new tab

3.5. Transcription factors analysis

This work found important protein kinases and transcription factors (TFs) linked to DEGs. These conclusions were drawn from their major contributions to the regulated network’s development. A regulatory network was built to develop regulatory complexity comprising kinases, TFs, and the transient proteins they were linked to.

To identify the main transcriptional regulators (TFs), the targeted genes found by ChIP-seq assays (ChEA) must be integrated in the first stage. These TFs were then mapped into networks of PPIs. The transcription factors (TFs) that have been found and the network produced by PPI are shown in Figs 6A and 6B. The main gene transcription factors found in this case are UBTF, NFIC, PPARG, AR, IRF8 and RCOR1, according to the hypergeometric p-value. Moreover, kinase entities, known as PPIs, were discovered and added to the network, indicating that they may play a regulatory role in the enlarged PPI interaction network. The highlighted kinase selections and the corresponding PPI network connections are shown in (Fig 5C–5D). The hypergeometric p-value analysis was also used to determine which dominant kinase was associated with these DEGs. Kinases such as ERK2, MAPK1, CSNK2A1, MAPK14, ERK1, GSK3B, CDK1, MAPK3, CDK4, and HIPK2 have been identified. On the other hand, the identify top 10 hub genes in GSE25496 data set key protein kinases and transcription factors were identified on the base of hypergeometric p-value analysis are display in S5 (A-B) Fig and (C-D). Tables S2 and S3 in S1 Fig the supplemental materials list every TF and kinase discovered and their scores.

Fig 6. A lifetime diagram displays every target gene (PPARG, AIF1, C1QA, SIGLEC1, MSR1, MS4A6A, PSEN2, ITPR2, PTX3, and ATF3).

Fig 6

Open in a new tab

When compared to the low expression of PPARG (black), the red survival curve for PPARG shows a higher survival rate (p < 0.0053). This trend holds true for all the genes based on statistical analysis using the Kaplan-Meier plotter.

Fig 5. (A) A bar graph illustrating the score and hypergeometric likelihood value for every detected element is displayed with a collection of anticipated transcription factors for the DEGs. (B) The components of the connected transcripts system and the interacting peptides are shown in a sphere-and-stick model. Gray nodes show the peptides that connect transcription factors, and pink nodes represent transcription factors. The degree of every node in the network is correlated with its size and quantity. (C) A bar graph displaying the kinase score (hypergeometric p-value), produced using the X2K Web tool, presents a ranking of the most likely results. (D) A visualization of the PPI networks corresponding to the most often identified kinases in differential gene expression.

Fig 5

Open in a new tab

3.6. Survival analysis of the hub genes

To examine the overall influence on survival of ten major hub genes chosen from both up and down regulated DEGs, we used KMPlot online tool. Among the identified genes investigated, only PTX3 and PPARG had a significant connection with decreased overall survival, with a hazard ratio (HR) of 1.63 and 1.47 respectively, in the group with greater expression levels. Survival analysis employing all hub genes and p-value predictions indicate that these genes significantly influence disease diagnosis and prognosis. These genes may be systematically prioritized in illness detection and therapeutic analysis. Through this KMPlot online tool estimated survival profiles for each hub gene are displayed in Fig 6.

3.7. Protein-drugs interactions

A drug repository database was consulted to choose medications based on the top 10 hub genes found (Table 5). This investigation used a manual search to find the right drugs for each target. The inquiry panels cover a wide range of medications, such as FDA-approved medications, experimental medications, and nutraceuticals. A manual screening of medications was done to find interactions with every hub gene-identified protein. For each of them, a total of 11 medications were found. PPARG was linked to two medications, whereas the other hub genes were each connected to a single drug. To guarantee the safety of AMG-131 and Elafibranor in clinical settings, their possible adverse effects should be carefully examined. While Elafibranor, a PPARα/γ dual agonist, has been associated in clinical trials with hepatotoxicity, gastrointestinal problems, and possible cardiovascular adverse effects, AMG-131, a PPARG antagonist, may cause metabolic disruptions, insulin resistance, and cardiovascular risks. Analyzing their toxicity, pharmacokinetics, and off-target effects requires a thorough ADMET analysis. The ADMET properties have been inferred via SWISS-ADME server, and the results are shown in Table S4 in S1 Fig.

Table 5. Drugs that target protein targets can be identified by analyzing the Drug Bank database.

S.NO Protein Uniprot ID Drug name Drug bank ID Groups
1. PPARG P37231

Q03181 		AMG-131

Elafibranor		 DB05490

DB05187		 Investigational

Investigational
2. AIF1 Q9BZX2 samarium (153Sm) lexidronam DB05273 Approved
3. C1QA Q9BZX2 Clemastine DB00283 Approved
4. SIGLEC1 P23975 Iobenguane DB06704 Approved
5. MSR1 Q96SW2 Thalidomide DB01041 Approved
6. MS4A6A P40227 Morphine DB00295 Approved
7. PSEN2 Q9UHC9 Cimetidine DB00501 Approved
8. ITPR2 Q6ZQN7 Saxagliptin DB06335 Approved
9. PTX3 P30968 Leuprolide DB00007 Approved
10. ATF3 P41143 Tapentadol DB06204 Approved
Open in a new tab

3.8. Molecular Docking of PPARG and its potential drug

A structure-based virtual screening of drug-like compounds was performed in molecular docking to find compounds with a higher affinity for the receptor enzyme active site. PyRx was used to dock inhibitors into the enzyme’s active site using the AutoDock Vina plugin for both compounds obtained during the computational screening [42]. Both the best inhibitors were determined to be compound (AMG-131) with the chemical formula [2,4-dichloro-N-(3,5-dichloro-4-quinolin-3-yloxyphenyl)benzenesulfonamide] and (Elafibranor) with the chemical formula [2-[2,6-dimethyl-4-[(E)-3-(4-methyl sulfanyl phenyl)-3-oxoprop-1-enyl]phenoxy]-2-methyl propanoic acid] Fig 7. The compounds binding affinity results in −7.4 kcal/mol and −7.2 kcal/mol, respectively. Results were analyzed for control inhibitor, i.e., (Rosiglitazone (S) which inferred −7.5 kcal/mol binding affinity with interactive residues Thr278, Arg371, Gln275, Glu243, Arg316 and His315 as shown in Fig 8C. According to Fig 8, the chemical was examined for interactions with almost identical active site residues. It was discovered that the compound quinolin-3-yloxyphenyl)benzenesulfonamide] ring and -(4-methyl sulfanyl phenyl)- extended toward the enzyme N-terminus. Alongside, the molecular docking analysis for others pharmacological targets elucidate the binding affinities of five compounds—AMG-131, Elafibranor, Rosiglitazone (S), Samarium Sm-153 Lexidronam, and Clemastine toward three receptors: 1FM6, 2g2b, and 1udw. AMG-131 demonstrates robust binding to all receptors, with the greatest affinity recorded for 1FM6 (−7.4 kcal/mol), followed by 1udw target (−7.1 kcal/mol) and 2g2b target (−6.9 kcal/mol). Elafibranor exhibits substantial binding, especially with 1FM6 (−7.2 kcal/mol), but significantly reduced affinities are seen for 2g2b (−5.8 kcal/mol) and 1udw (−5.6 kcal/mol). The positive control molecule, Rosiglitazone (S), has the highest binding affinity with 1FM6 (−7.5 kcal/mol) and maintains consistent interactions with target 2g2b (−6.1 kcal/mol) and 1udw (−6.1 kcal/mol). Samarium Sm-153 Lexidronam and Clemastine exhibit decreased receptor binding, with affinities between −5.2 kcal/mol and −6.9 kcal/mol respectively as shown in Table S5 in S1 Fig. These findings emphasize AMG-131 and Rosiglitazone (S) as viable options, especially for targeting the 1FM6 receptor, while revealing receptor-specific characteristics that might inform future optimization techniques. Alongside, a negative control drug) was run against the PPARG target which resulted in a binding affinity of (−5.2 kcal/mol) against 1FM6. (−4.5 kcal/mol) against 2g2b and (−4.3 kcal/mol) in response to 1udw respectively with low binding interactions. These results confirm the compounds’ reliability in preventing the substrate from reaching the active site, preventing the formation of the enzyme-substrate complex and the product. The desired configuration of the molecule involved interactions between the Arg76, Leu28leu Leu118, a residue of the loop region from the central domain against both the compounds inferred strong hydrogen bonds formation and other bonds inside the active domain as shown in Fig 8A-8B and 8C. The distance of hydrogen bonds among residue Arg76 with ligand atom (Fig 8A) is 2.3 Å, 3.4 Å (Fig 8B) along with the other residues whose average distance score results in 3.6 Å shows a high binding affinity.

Fig 7. Best hit molecules against PPARG protein target.

Fig 7

Open in a new tab

Fig 8. (A) Depicting the active pocket with interactive residues at the N-terminal end against both the inhibitors. (B) showing the linkages among the receptor and active compounds with benzene rings presenting strong hydrogen bonds, pi-sigma and Alkyl bonds. (C) Presenting the control inhibitor against the target protein with binding interactions in 3D and 2D analysis with strong hydrogen bonds and other linkages.

Fig 8

Open in a new tab

According to literature in ligand-displacement direct binding assays, the highly potent PPARG ligand 2,4-dichloro-N-(3,5-dichloro-4-(quinolin-3-yloxy)phenyl)benzenesulfonamide has a Ki of 10 nM, which is enough to push rosiglitazone out of the ligand binding pocket [43]. This compound is much more effective than rosiglitazone in cell-based transcriptional activation tests, with an EC50 value of 4 nM and a peak transcriptional activation of reporter genes of about 30%. Cell-based functional tests have demonstrated distinct coregulatory recruitment characteristics of comparison to complete agonists like rosiglitazone [44]. Whereas the compound Elafibranor is an agonist of both peroxisome proliferator-activated receptor-α and peroxisome proliferator-activated receptor-δ. have just been granted FDA expedited clearance after phase 3 studies that showed significant increases in ALP, biochemical response rates, and pruritus alleviation. It can regulate lipid metabolism, enhance glucose homeostasis, and have anti-inflammatory properties [45]. Elafibranor long-term safety and impact on clinical outcomes were assessed in a phase 2 study. Additionally, elafibranor medication was linked to a significant reduction in immunologic and inflammatory indicators, such as circulating levels of the bile acid precursor 7α-hydroxy-4-cholesten-3-one (C4), and immunoglobulin M (IgM) [46].

3.9. MD simulation

The Peroxisome proliferator-activated receptor gamma enzyme (PPARG) has several static configurations that have been proposed, either by itself or in conjunction with ligand; it is still necessary to understand the dynamics of the enzyme in an aqueous milieu to properly function. A 200-ns timeline was used to study the dynamics of the protein in combination with the best inhibitor. This makes it possible to develop new effective inhibitors against the enzyme and provides a thorough understanding of the structural changes made to the protein in the presence of ligand. The stability of the protein in complex with ligand was examined during the simulation period using four statistical parameters: Radius of Gyration (Rg), Root Mean Square Fluctuation (RMSF), and RMSD (Fig 9A). The distances between the backbone atoms of overlaid proteins are averaged and expressed as RMSD. The RMSD of the Cα atoms suggests that the enzyme-inhibitor combination is stable throughout the time intervals. The projected mean RMSD value was 2.95 Å for the AMG-131 complex system and 2.93 Å against the Elafibranor complex system, with a maximum RMSD value of 3.7 Å at 100 ns and 4.1 Å at 166 ns, respectively. RMSF values were tracked together with the convergence values for both systems to uncover the structural mobility for every system residue. It is seen that the enzyme residues in confined form remained extremely stable in both systems. The mean root mean square function (RMSF) values for the complex systems were 1.3Å for AMG-131 and Elafibranor complex systems, respectively. Herein, the maximum RMSF value recorded for both systems is 7.5Å at residue 57 and 14.1 Å at the start of the residue. Most residues of the docked protein exhibit RMSF values <2 Å, as Fig 9B clearly illustrates. This indicates the PPARG enzyme’s extreme stability even further. Finally, to gain a deeper knowledge of the systems, the equilibrium conformation of the systems was examined. This was accomplished by figuring out Rg, a measure of how compact a protein structure is. During a simulation run, changes in the Rg measurements affect how compact the protein is with time intervals as shown in Fig 9C. Protein amino acids are packed loosely in higher Rg values but tightly in lower Rg values. With a mean value of 19.3 Å for complex AMG-131, a maximum value of 20 Å was recorded at 194 ns and a mean value of 18.9 Å for complex Elafibranor was recorded with a maximum value of 19.5Å at 134 ns. The Rg values for both systems presented here confirm the protein’s stability in docked form and are comparable to the RMSD values.

Fig 9. Depicting the molecular dynamics of complexes obtained during the system biology approach.

Fig 9

Open in a new tab

(A) Depicting the RMSD of both complex systems for 200 ns simulation intervals. (B) The flexibility and thermal effects on residues during the simulation time frame of 200 ns for both complex systems were presented with RMSF analysis, and (C) the rigidity of carbon alpha atoms with radius of gyration analysis was illustrated for both complex systems.

Root mean square deviation (RMSD) and root mean square fluctuation (RMSF) analyses were used in our 100 ns molecular dynamics simulations to quantify the relative stability and binding kinetics of the AMG-131 and Elafibranor complexes (Fig 9). A stiff and well-bound conformation was shown by the AMG-131 complex’s lower and more stable backbone and ligand RMSD profiles (Figs 9A, 9B), which were further corroborated by the RMSF plot decreased fluctuations in binding-site residues (Fig 9C). The ligand’s limited angular displacement within the pocket over time confirmed its spatial stability (Fig 10). On the other hand, the Elafibranor complex showed increased residue flexibility (Fig 9C) and higher backbone and ligand RMSD values with larger variability (Figs 9A, 9B), which is in line with its notable angular reorientations seen along the trajectory (Fig 10). Together, these findings imply that, in contrast to the more dynamic and mobile Elafibranor-bound state, AMG-131 forms a more stable and conformationally restricted complex.

Fig 10. Depicting the complex systems after 0 ns and 100 ns time intervals with minor angular displacement of ligand inside the pocket throughout the simulations.

Fig 10

Open in a new tab

4. Discussion

Comprehensive investigation of molecular changes linked to disease has been made possible by developments in digital and genomic technologies, such as SNP analysis, GWAS, and microarray-based expression profiling. Through multi-level omics, such as transcriptomics, proteomics, metabolomics, and genomes, these techniques facilitate the identification of biomarkers and enhance comprehension of disease causes. [16,47]. In this study, protein–protein interaction (PPI) networks were constructed to highlight important regulators after differentially expressed genes (DEGs) were identified using microarray datasets (GSE3624 and GSE25496). Although instructive, both datasets’ small sample sizes lower statistical power and raise the possibility of false positives or false negatives, highlighting the need for careful interpretation and further validation with bigger cohorts. PPI network analysis is frequently used to understand metabolic processes, find possible treatment targets, and clarify disease causes [48]. The structural relationships between several proteins greatly influence numerous biological processes, which might differ between healthy and pathological states [48,49]. The identification of genes that are expressed differently in disease situations than in healthy ones, microarray gene expression data analysis helps identify targets for new medication development. Molecular network connections improve the precision and consistency of identifying illness-associated biomarkers [50]. Prior studies have indicated the significance of these analyses in forecasting central nodes and their vital roles in a range of illnesses [51,52].

Regarding PLOSL patients, the prognosis for those in the early stages is far better than that for those in the later stages. Early-stage illness patients frequently experience excellent cure rates while receiving chemotherapy, radiation therapy, surgery, or a combination of these therapies. On the other hand, those who have advanced-stage PLOSL have more difficulties, mainly because the illness is aggressive, and a solution is sometimes unattainable [53]. This study used microarray data to find DEGs in PLOSL. Ten hub genes were shown to be potential contributors to PLOSL disease by PPI network and enrichment analysis, with PPARG appearing as the most promising candidate. Prior research demonstrates that PPARG is involved in processes related to the development of PLOSL, such as inflammation, lipid regulation, and cell-proliferative activity [54]. It has been demonstrated that PPARG has a major influence on PLOSL formation [55]. PPARG’s role in disease-related processes, such as increased cell proliferation, made it a viable target for intervention [56]. In a recent study, it has been revealed that microglial PPARG signaling promotes the wound healing process by modulating transcriptional events, such as limiting chronic microglial activation. This explains the causal role of PPARG in downstream functional decline and in regulating neuroinflammation [57]. Additionally, preclinical findings suggest the suppression of NF-κB/inflammasome signaling pathways and microglia shift to anti-inflammatory phenotypes, illustrating the PPARG as a neuroinflammation hub [58]. The roles of the other identified targets are critical to the evolution of PLOSL. Their relevance was highlighted by enrichment analysis, subnetwork building, and the incorporation of all hub genes inside these subnetworks. A review of KEGG pathways, molecular activities, cellular constituents, and biological processes clarified the functions of the genes in several interrelated pathways. Identifying transcription factors and their roles within the expanded network enabled determining their regulatory functions. Previous studies have noted the importance of ECM–receptor interaction and Notch signaling pathways in PLOSL, also included in our KEGG pathway analysis. Additionally, our research indicates significant alterations in immune-related pathways in PLOSL patients [53].

However, several other hub genes that have been found, like AIF1 and C1QA, also seem to have biological significance. C1QA is essential for immune-mediated clearance and synaptic remodeling, whereas AIF1 is a microglial activation marker associated with phagocytic regulation and neuroinflammation. When taken as a whole, these genes demonstrate the significant role that dysregulated immunological signaling, microglial dysfunction, and compromised efferocytosis play in PLOSL. Their dispersion among biological processes, molecular activities, and KEGG pathways provides more evidence that the disease is complex rather than dependent on a single dominant driver.

Molecular dynamics (MD) simulations, docking, and structural modeling were used concurrently to evaluate PPARG as a possible therapeutic target. Docking was used to screen potential FDA-approved drugs, and then MD simulations were used to assess receptor–ligand stability. Although AMG-131 and Elafibranor showed good in-silico binding characteristics, these results are still in the early stages. Crucially, AMG-131 is still an early-stage research chemical without therapeutic licensure, and Elafibranor’s clinical development has been halted due to unsatisfactory trials. As a result, none of the discovered compounds are currently thought to be clinically feasible for PLOSL, and thorough in-vitro and in-vivo validation of computational predictions is necessary. Even though our computational investigations showed that AMG-131 and Elafibranor have favorable binding interactions with PPARG, it is crucial to remember that neither drug is now therapeutically viable. The development of Elafibranor, a dual PPARα/γ agonist, was stopped after it failed to meet primary endpoints in clinical trials. There is currently no authorized therapeutic indication for AMG-131, making it an early-stage investigative chemical. Therefore, before any evaluation of therapeutic relevance, the results given here should be carefully understood as preliminary in-silico discoveries that require considerable preclinical and experimental validation.

5. Conclusions

To find possible molecular contributors and prospective drugs related to PLOSL, this study used differential expression analysis, PPI network design, functional enrichment, and structure-based computational modeling. PPARG and many other hub genes, such as AIF1 and C1QA, were identified by the analysis as potential important regulators of immune-related pathways and microglial dysfunction. The molecular processes that may explain PLOSL disease are first revealed by these results. AMG-131 and Elafibranor have preferential binding to PPARG, according to computational docking and molecular dynamics simulations; nonetheless, these findings should be interpreted cautiously. With AMG-131 still in the experimental stage and Elafibranor no longer in clinical research, both substances are currently unvalidated in the context of PLOSL. Consequently, treatment preparedness is not indicated by the interaction predictions shown here. All things considered, this work provides an initial framework for researching PLOSL that generates hypotheses. The results are exploratory and computational in nature; further in-vitro, in-vivo, and clinical validation will be required to ascertain the actual biological significance of the identified genes and chemicals. To get these candidates closer to any possible therapeutic use, more extensive datasets and experimental validation must be included in future studies.

Finding new pharmacological targets that contribute to the onset and development of PLOSL is crucial, in addition to the biomarkers already used for the condition, which contain mutations that result in therapy resistance. The statistical strength of the results is limited because the datasets employed in this investigation were based on very small sample sizes. Even though our computational analysis suggested PPARG and other hub genes as possible targets for treatment, these findings should be considered preliminary. To establish the biological significance of the suggested indicators and to further assess the therapeutic potential of AMG-131 and Elafibranor, future studies utilizing larger patient cohorts and rigorous experimental validation in both in vitro and in vivo models are crucial.

This research aims to identify novel biomarkers linked to PLOSL that may be used to control the condition’s occurrence worldwide. The two therapeutic drugs, AMG-131 and Elafibranor, were chosen for binding study against the target gene and verified using molecular docking methods based on encouraging scores along with the control inhibtor. This study will help us better understand how drugs resist their effects and pave the way for developing novel therapeutics and diagnostics. The drug’s mechanical behavior and mechanism of action toward the target biomarker gene PPARG were revealed by MD investigations. The drugs under study bind tightly to the active site and maintain this binding throughout the simulation period. This strategy for biomarker discovery delivers more accuracy and cost-effectiveness than conventional experimental approaches by utilizing artificial intelligence and computational methodologies, eventually enhancing PLOSL research. The predicted compounds can be experimentally tested using different in vitro approaches; binding assays, enzymatic assays, cell-based functional assays, and phenotypic screening (Jennings, 2015). The drug can be further validated using in vivo methods in animal models such as like rats and mice.

Supporting information

S1 Fig. Prior to normalization, the box plot, density plot, and mean expression plot are shown in (A), (B), and (C); following normalization, the same plots are shown in (D), (E), and (F). Using the Network Analyst Tool, variance-stabilizing normalization successfully removed noise from the data. The next methods used to analyze the data were Principal Component Analysis and the Quality Control Examination.

(DOCX)

pone.0343274.s001.docx (1.3MB, docx)

Data Availability

All relevant data are within the manuscript and its Supporting Information files.

Funding Statement

The author(s) received no specific funding for this work.

References

  • 1.Adam MP, Feldman J, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A. Polycystic lipomembranous osteodysplasia with sclerosing leukoencephalopathy. GeneReviews®. Seattle (WA): University of Washington; 1993. http://www.ncbi.nlm.nih.gov/books/NBK1197 [PubMed] [Google Scholar]
  • 2.Schrader H, Lund RT, Halvorsen TB, Møen KM, Bergan U. Polycystic lipomembranous osteodysplasia with sclerosing leukoencephalopathy. Tidsskr Nor Laegeforen. 2000;120(21):2520–3. [PubMed] [Google Scholar]
  • 3.Xing J, Titus AR, Humphrey MB. The TREM2-DAP12 signaling pathway in Nasu-Hakola disease: a molecular genetics perspective. Res Rep Biochem. 2015;5:89–100. doi: 10.2147/RRBC.S58057 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Paloneva J, Manninen T, Christman G, Hovanes K, Mandelin J, Adolfsson R, et al. Mutations in two genes encoding different subunits of a receptor signaling complex result in an identical disease phenotype. Am J Hum Genet. 2002;71(3):656–62. doi: 10.1086/342259 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Grodzka O, Andrzejczak-Sobocińska A, Procyk G, Majewski G, Słyk S, Motyl R, et al. A mild type of Nasu-Hakola disease - a case of a woman with presenile dementia and cystic bone lesions. Arch Med Sci. 2023;19(5):1589–92. doi: 10.5114/aoms/169476 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Myers S, Baker A. Drug discovery--an operating model for a new era. Nat Biotechnol. 2001;19(8):727–30. doi: 10.1038/90765 [DOI] [PubMed] [Google Scholar]
  • 7.DiMasi JA, Hansen RW, Grabowski HG. The price of innovation: new estimates of drug development costs. J Health Econ. 2003;22(2):151–85. doi: 10.1016/S0167-6296(02)00126-1 [DOI] [PubMed] [Google Scholar]
  • 8.Lobanov V. Using artificial neural networks to drive virtual screening of combinatorial libraries. Drug Discovery Today: BIOSILICO. 2004;2(4):149–56. doi: 10.1016/s1741-8364(04)02402-3 [DOI] [Google Scholar]
  • 9.Braga RC, Alves VM, Silva AC, Nascimento MN, Silva FC, Liao LM, et al. Virtual screening strategies in medicinal chemistry: the state of the art and current challenges. Curr Top Med Chem. 2014;14(16):1899–912. doi: 10.2174/1568026614666140929120749 [DOI] [PubMed] [Google Scholar]
  • 10.Grover S, Apushkin MA, Fishman GA. Topical dorzolamide for the treatment of cystoid macular edema in patients with retinitis pigmentosa. Am J Ophthalmol. 2006;141(5):850–8. doi: 10.1016/j.ajo.2005.12.030 [DOI] [PubMed] [Google Scholar]
  • 11.Beddell CR, Goodford PJ, Norrington FE, Wilkinson S, Wootton R. Compounds designed to fit a site of known structure in human haemoglobin. Br J Pharmacol. 1976;57(2):201–9. doi: 10.1111/j.1476-5381.1976.tb07468.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JPA, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1-34. doi: 10.1016/j.jclinepi.2009.06.006 [DOI] [PubMed] [Google Scholar]
  • 13.Kiialainen A, Veckman V, Saharinen J, Paloneva J, Gentile M, Hakola P, et al. Transcript profiles of dendritic cells of PLOSL patients link demyelinating CNS disorders with abnormalities in pathways of actin bundling and immune response. J Mol Med (Berl). 2007;85(9):971–83. doi: 10.1007/s00109-007-0191-4 [DOI] [PubMed] [Google Scholar]
  • 14.Xia J, Gill EE, Hancock REW. NetworkAnalyst for statistical, visual and network-based meta-analysis of gene expression data. Nat Protoc. 2015;10(6):823–44. doi: 10.1038/nprot.2015.052 [DOI] [PubMed] [Google Scholar]
  • 15.Stacklies W, Redestig H, Scholz M, Walther D, Selbig J. pcaMethods--a bioconductor package providing PCA methods for incomplete data. Bioinformatics. 2007;23(9):1164–7. doi: 10.1093/bioinformatics/btm069 [DOI] [PubMed] [Google Scholar]
  • 16.Khan A, Rehman Z, Hashmi HF, Khan AA, Junaid M, Sayaf AM, et al. An Integrated Systems Biology and Network-Based Approaches to Identify Novel Biomarkers in Breast Cancer Cell Lines Using Gene Expression Data. Interdiscip Sci. 2020;12(2):155–68. doi: 10.1007/s12539-020-00360-0 [DOI] [PubMed] [Google Scholar]
  • 17.Davis J, Gentleman R, Carey V, Huber W, Irizarry R, Dudoit S. Journal of the American Statistical Association. 2007;102:388–9. [Google Scholar]
  • 18.Jiao X, Sherman BT, Huang DW, Stephens R, Baseler MW, Lane HC, et al. DAVID-WS: a stateful web service to facilitate gene/protein list analysis. Bioinformatics. 2012;28(13):1805–6. doi: 10.1093/bioinformatics/bts251 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet. 2000;25(1):25–9. doi: 10.1038/75556 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Szklarczyk D, Franceschini A, Wyder S, Forslund K, Heller D, Huerta-Cepas J, et al. STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res. 2015;43(Database issue):D447-52. doi: 10.1093/nar/gku1003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Chen Y-C, Hsiao C-C, Chen K-D, Hung Y-C, Wu C-Y, Lie C-H, et al. Peripheral immune cell gene expression changes in advanced non-small cell lung cancer patients treated with first line combination chemotherapy. PLoS One. 2013;8(2):e57053. doi: 10.1371/journal.pone.0057053 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–504. doi: 10.1101/gr.1239303 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Xia J, Benner MJ, Hancock REW. NetworkAnalyst--integrative approaches for protein-protein interaction network analysis and visual exploration. Nucleic Acids Res. 2014;42(Web Server issue):W167-74. doi: 10.1093/nar/gku443 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Ma L, Huang Y, Zhu W, Zhou S, Zhou J, Zeng F, et al. An integrated analysis of miRNA and mRNA expressions in non-small cell lung cancers. PLoS One. 2011;6(10):e26502. doi: 10.1371/journal.pone.0026502 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Ghafouri-Fard S, Safarzadeh A, Taheri M, Jamali E. Identification of diagnostic biomarkers via weighted correlation network analysis in colorectal cancer using a system biology approach. Sci Rep. 2023;13(1). doi: 10.1038/s41598-023-40953-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Pan J, Liu B, Dai Z. The role of a lung vascular endothelium enriched gene TMEM100. Biomedicines. 2023;11(3):937. doi: 10.3390/biomedicines11030937 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Brandes U, Eiglsperger M, Herman I, Himsolt M, Marshall MS. GraphML Progress Report Structural Layer Proposal. Lecture Notes in Computer Science. Springer Berlin Heidelberg; 2002. 501–12. doi: 10.1007/3-540-45848-4_59 [DOI] [Google Scholar]
  • 28.Menyhárt O, Nagy Á, Győrffy B. Determining consistent prognostic biomarkers of overall survival and vascular invasion in hepatocellular carcinoma. R Soc Open Sci. 2018;5(12):181006. doi: 10.1098/rsos.181006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Wishart DS, Feunang YD, Guo AC, Lo EJ, Marcu A, Grant JR, et al. DrugBank 5.0: a major update to the DrugBank database for 2018. Nucleic Acids Res. 2018;46(D1):D1074–82. doi: 10.1093/nar/gkx1037 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Dallakyan S, Olson AJ. Small-molecule library screening by docking with PyRx. Chemical biology. Springer. 2015. p. 243–50. [DOI] [PubMed] [Google Scholar]
  • 31.Gampe RT Jr, Montana VG, Lambert MH, Miller AB, Bledsoe RK, Milburn MV, et al. Asymmetry in the PPARgamma/RXRalpha crystal structure reveals the molecular basis of heterodimerization among nuclear receptors. Mol Cell. 2000;5(3):545–55. doi: 10.1016/s1097-2765(00)80448-7 [DOI] [PubMed] [Google Scholar]
  • 32.Case DA, Betz RM, Cerutti DS, Cheatham T, Darden T, Duke RE. Amber 16. University of California, San Francisco: University of California, San Francisco; 2016. doi: 10.13140/rg.2.2.27958.70729 [DOI] [Google Scholar]
  • 33.Wang J, Wang W, Kollman PA, Case DA. Antechamber: an accessory software package for molecular mechanical calculations. J Am Chem Soc. 2001;222:U403. [Google Scholar]
  • 34.Sprenger KG, Jaeger VW, Pfaendtner J. The general AMBER force field (GAFF) can accurately predict thermodynamic and transport properties of many ionic liquids. J Phys Chem B. 2015;119(18):5882–95. doi: 10.1021/acs.jpcb.5b00689 [DOI] [PubMed] [Google Scholar]
  • 35.Case DA, Babin V, Berryman JT, Betz RM, Cai Q, Cerutti DS. The FF14SB force field. Amber. 2014;14:29–31. [Google Scholar]
  • 36.Schafmeister C, Ross WS, Romanovski V. The leap module of AMBER. University of California; 1995. [Google Scholar]
  • 37.Brice AR, Dominy BN. Examining electrostatic influences on base-flipping: a comparison of tip3p and gb solvent models. Commun comput phys. 2013;13(1):223–37. doi: 10.4208/cicp.210711.121011s [DOI] [Google Scholar]
  • 38.Krautler V, van Gunsteren WF, Hunenberger PH. A fast SHAKE algorithm to solve distance constraint equations for small molecules in molecular dynamics simulations. J Comput Chem. 2001;22(5):501–8. doi: 10.1002/1096-987x(20010415)22:5<501::aid-jcc1021>3.0.co;2-v [DOI] [Google Scholar]
  • 39.Kuzmanic A, Zagrovic B. Determination of ensemble-average pairwise root mean-square deviation from experimental B-factors. Biophys J. 2010;98(5):861–71. doi: 10.1016/j.bpj.2009.11.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Berendsen HJC, Postma JPM, van Gunsteren WF, DiNola A, Haak JR. Molecular dynamics with coupling to an external bath. The Journal of Chemical Physics. 1984;81(8):3684–90. doi: 10.1063/1.448118 [DOI] [Google Scholar]
  • 41.Roe DR, Cheatham TE 3rd. PTRAJ and CPPTRAJ: Software for processing and analysis of molecular dynamics trajectory data. J Chem Theory Comput. 2013;9(7):3084–95. doi: 10.1021/ct400341p [DOI] [PubMed] [Google Scholar]
  • 42.Zhao S, Ni F, Qiu T, Wolff JT, Tsai S-C, Luo R. Molecular basis for polyketide ketoreductase-substrate interactions. Int J Mol Sci. 2020;21(20):7562. doi: 10.3390/ijms21207562 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Motani A, Wang Z, Weiszmann J, McGee LR, Lee G, Liu Q, et al. INT131: a selective modulator of PPAR gamma. J Mol Biol. 2009;386(5):1301–11. doi: 10.1016/j.jmb.2009.01.025 [DOI] [PubMed] [Google Scholar]
  • 44.Frkic RL, He Y, Rodriguez BB, Chang MR, Kuruvilla D, Ciesla A, et al. Structure-Activity Relationship of 2,4-Dichloro-N-(3,5-dichloro-4-(quinolin-3-yloxy)phenyl)benzenesulfonamide (INT131) Analogs for PPARγ-Targeted Antidiabetics. J Med Chem. 2017;60(11):4584–93. doi: 10.1021/acs.jmedchem.6b01727 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Jamal F, Elshaer A, Alatout MH, Odeh NB, Youssef A, Abo Abdullah H, et al. Second-line therapies in primary biliary cholangitis: a comparative review of obeticholic acid, fibrates, seladelpar, and elafibranor. Biomedicines. 2025;13(10):2335. doi: 10.3390/biomedicines13102335 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Schattenberg JM, Pares A, Kowdley KV, Heneghan MA, Caldwell S, Pratt D, et al. A randomized placebo-controlled trial of elafibranor in patients with primary biliary cholangitis and incomplete response to UDCA. J Hepatol. 2021;74(6):1344–54. doi: 10.1016/j.jhep.2021.01.013 [DOI] [PubMed] [Google Scholar]
  • 47.Akcay IM, Katrinli S, Ozdil K, Doganay GD, Doganay L. Host genetic factors affecting hepatitis B infection outcomes: Insights from genome-wide association studies. World J Gastroenterol. 2018;24(30):3347–60. doi: 10.3748/wjg.v24.i30.3347 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Hu B, Chang X, Liu X. Predicting functional modules of liver cancer based on differential network analysis. Interdiscip Sci. 2019;11(4):636–44. doi: 10.1007/s12539-018-0314-3 [DOI] [PubMed] [Google Scholar]
  • 49.Ni Q-S, Wang Z-Z, Li G-G, Wang G-Y, Zhao Y-J. Effect of the quality of the interaction data on predicting protein function from protein-protein interactions. Interdiscip Sci. 2009;1(1):40–5. doi: 10.1007/s12539-008-0015-4 [DOI] [PubMed] [Google Scholar]
  • 50.Ma R, Wang C, Wang J, Wang D, Xu J. miRNA-mRNA interaction network in non-small cell lung cancer. Interdiscip Sci. 2016;8(3):209–19. doi: 10.1007/s12539-015-0117-8 [DOI] [PubMed] [Google Scholar]
  • 51.Nayarisseri A, Yadav M, Wishard R. Computational evaluation of new homologous down regulators of Translationally Controlled Tumor Protein (TCTP) targeted for tumor reversion. Interdiscip Sci. 2013;5(4):274–9. doi: 10.1007/s12539-013-0183-8 [DOI] [PubMed] [Google Scholar]
  • 52.Khan A, Ali A, Junaid M, Liu C, Kaushik AC, Cho WCS, et al. Identification of novel drug targets for diamond-blackfan anemia based on RPS19 gene mutation using protein-protein interaction network. BMC Syst Biol. 2018;12(Suppl 4):39. doi: 10.1186/s12918-018-0563-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Henson JW. Treatment of glioblastoma multiforme: a new standard. Arch Neurol. 2006;63(3):337–41. doi: 10.1001/archneur.63.3.337 [DOI] [PubMed] [Google Scholar]
  • 54.Rao JS, Yamamoto M, Mohaman S, Gokaslan ZL, Fuller GN, Stetler-Stevenson WG, et al. Expression and localization of 92 kDa type IV collagenase/gelatinase B (MMP-9) in human gliomas. Clin Exp Metastasis. 1996;14(1):12–8. doi: 10.1007/BF00157681 [DOI] [PubMed] [Google Scholar]
  • 55.Park SY, Piao Y, Jeong KJ, Dong J, de Groot JF. Periostin (POSTN) Regulates Tumor Resistance to Antiangiogenic Therapy in Glioma Models. Mol Cancer Ther. 2016;15(9):2187–97. doi: 10.1158/1535-7163.MCT-15-0427 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Wang Y, Sun Q, Geng R, Liu H, Yuan F, Xu Y, et al. Notch intracellular domain regulates glioblastoma proliferation through the Notch1 signaling pathway. Oncol Lett. 2021;21(4):303. doi: 10.3892/ol.2021.12564 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Zeng Q, Zhou T-T, Huang W-J, Huang X-T, Huang L, Zhang X-H, et al. Asarinin attenuates bleomycin-induced pulmonary fibrosis by activating PPARγ. Sci Rep. 2023;13(1):14706. doi: 10.1038/s41598-023-41933-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Fang M, Yu Q, Ou J, Lou J, Zhu J, Lin Z. The Neuroprotective Mechanisms of PPAR-γ: Inhibition of Microglia-Mediated Neuroinflammation and Oxidative Stress in a Neonatal Mouse Model of Hypoxic-Ischemic White Matter Injury. CNS Neurosci Ther. 2024;30(11):e70081. doi: 10.1111/cns.70081 [DOI] [PMC free article] [PubMed] [Google Scholar]

Decision Letter 0

Md Zubbair Malik

12 Nov 2025

-->PONE-D-25-34800-->-->System Biology and Network-Based Approach to Identify the Therapeutic Signatures and Potential Inhibitors Against Polycystic Lipomembranous Osteodysplasia with Sclerosing Leukoencephalopathy-->-->PLOS ONE

Dear Dr. Alaa Abdulaziz,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.-->

Indicate which changes you require for acceptance versus which changes you recommend.

Please submit your revised manuscript by  Dec 27 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org . When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please note that PLOS One has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, we expect all author-generated code to be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse.

3. We note that you have indicated that there are restrictions to data sharing for this study. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

Before we proceed with your manuscript, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., a Research Ethics Committee or Institutional Review Board, etc.). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of recommended repositories, please see

https://journals.plos.org/plosone/s/recommended-repositories. You also have the option of uploading the data as Supporting Information files, but we would recommend depositing data directly to a data repository if possible.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

5. If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

-->Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. -->

Reviewer #1: Yes

Reviewer #2: Yes

**********

-->2. Has the statistical analysis been performed appropriately and rigorously? -->

Reviewer #1: N/A

Reviewer #2: Yes

**********

-->3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.-->

Reviewer #1: Yes

Reviewer #2: No

**********

-->4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.-->

Reviewer #1: Yes

Reviewer #2: Yes

**********

-->5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)-->

Reviewer #1: The authors have conducted a good work addressing a relevant scientific problem and have presented their analyses clearly. A few constructive suggestions have been provided to enhance clarity and strengthen the overall quality of the manuscript.

1. The study uses very small sample sizes, which limits statistical confidence. The authors should acknowledge the limitation clearly in the Discussion and Conclusion.

2. The authors should strengthen the rationale for selecting PPARG as the primary hub gene. Ranking by degree score alone is not sufficient; additional biological justification and supporting literature linking PPARG specifically to PLOSL pathology should be provided.

3. Survival analysis appears inappropriate because no PLOSL clinical survival dataset exists. The analysis should be removed or explicitly clarified as exploratory and non-disease-specific.

4. The authors should reduce repetitive content, particularly in the Introduction and Discussion, to improve clarity and maintain a concise narrative.

5. The PCA and heatmap figures require improved resolution and clearer labeling to allow interpretability.

6. Figures should include clear color legends and larger, readable font sizes. Several elements, particularly those in white text, are difficult to interpret and should be improved for better visibility.

7. The KEGG and GO enrichment results are listed but not fully interpreted. Please explain how identified pathways relate mechanistically to PLOSL progression, especially microglial and osteoclast dysfunction.

8. The text states that PPARG is associated with disease progression but does not cite primary mechanistic literature. Please add supportive references.

9. The molecular docking results indicate moderate binding affinity, so the therapeutic relevance should be discussed cautiously. While the MD simulation plots (RMSD, RMSF, and Rg) are provided, the interpretation of these results should be expanded to clearly explain the structural stability and conformational behavior of the ligand–protein complexes.

10. The chemical structures of AMG-131 and Elafibranor should be included as figures to support clarity.

11. The authors should also acknowledge the discontinued clinical trial status of Elafibranor and the experimental development stage of AMG-131 to avoid implying that these compounds are clinically viable options at this stage.

12. The manuscript uses long sentences that reduce readability. Grammar, sentence structure, and transitions require careful language editing.

13. The Conclusion section should be moderated; currently it implies therapeutic development readiness. Please emphasize that experimental validation is still required.

14. The authors may consider briefly discussing alternative hub candidates identified (e.g., C1QA, AIF1) to balance interpretation and avoid appearing to pre-select PPARG.

15. The study would benefit from a short limitations paragraph summarizing dataset size, lack of clinical validation, and computational-only nature.

16. The Introduction provides general background on PLOSL but could be more concise. Several paragraphs repeat known disease characteristics. Reducing redundancy will improve focus.

17. The Introduction should state a clear research hypothesis and specific study objectives at the end. Currently, the transition into the aim of the study is broad and lacks direct specificity.

18. Authors are advised to incorporate more recent literature (within the past 5 years) on PPARG signaling in neuroinflammation and osteoclast regulation to better support the rationale for selecting PPARG as the hub gene in this study.

19. Authors are advised to justify the selection of the specific GEO datasets used in this study, particularly regarding sample relevance and availability. Additionally, the differential expression threshold should be clarified, as the stated FDR < 2 is not standard practice; please correct or explain the appropriate adjusted p-value cut-off for clarity and consistency.

20. The PPI network analysis describes the construction process, but network confidence score choices (0.4 or 0.75) vary in different parts of the text. Standardize and justify the confidence threshold.

21. The CytoHubba ranking selection uses degree only. Consider including at least one additional centrality parameter (e.g., closeness or betweenness) to strengthen hub gene selection.

22. The volcano plot and heatmap would benefit from labeling the key differentially expressed genes to enhance interpretability and allow readers to easily identify the most significant up- and down-regulated genes.

23. The GO/KEGG tables list terms effectively, but narrative interpretation is brief. Highlight the most biologically relevant pathways in the context of PLOSL pathogenesis.

24. The presentation of docking results compares binding scores, but no comparison is made with known strong PPARG ligands beyond Rosiglitazone. Including known ligand reference values would contextualize the strength of interactions.

25. The Results section occasionally shifts into speculative statements (e.g., therapeutic implications) that belong in Discussion. Please separate objective findings from interpretation.

Reviewer #2: The manuscript by Bokhari et al. presents a comprehensive computational study aimed at identifying key hub genes, pathways, and potential drug repurposing candidates for the rare and severe Nasu-Hakola disease (PLOSL). The authors employ a systems biology workflow, including differential gene expression analysis, protein-protein interaction (PPI) network construction, survival analysis, and sophisticated computational chemistry techniques like molecular docking and molecular dynamics (MD) simulations. The study's main claims are the identification of PPARG and AIF1 as central hub genes in PLOSL pathology and the proposal of AMG-131 and Elafibranor as promising therapeutic agents targeting PPARG. The significance of this work lies in its attempt to address a major therapeutic gap for a devastating rare disease using a cost-effective, bioinformatics-driven drug repurposing approach. The application of a 200 ns MD simulation to validate docking results adds considerable strength to the stability claims of the proposed drug-target complexes.

Well for authors must note that PLOS One encourages submissions across the methodological spectrum, but for a study making specific claims about therapeutic potential and biomarker identification, some form of experimental corroboration is increasingly expected to justify publication.

It would be amicable if Authors can provide any experimental data, however preliminary, to validate your key computational findings? As correctly noted in the manuscript's conclusion, a direct binding assay (e.g., Surface Plasmon Resonance - SPR, or Isothermal Titration Calorimetry - ITC) to demonstrate the physical interaction between PPARG and the proposed drugs (AMG-131, Elafibranor) would be highly impactful. For instance:

(i) A cell-based functional assay in a relevant cell line (e.g., microglial or osteoclast precursors) could test whether these compounds can modulate PPARG activity or reverse a PLOSL-associated phenotype.

(ii) qPCR or Western Blot analysis to validate the differential expression of the top hub genes (PPARG, AIF1, C1QA, etc.) in a cellular or animal model of PLOSL would significantly strengthen the initial bioinformatics findings.

(iii) The manuscript would be strengthened by explicitly stating the distance in angströms (Å) for the key hydrogen bonds, especially the one with Arg76 as shown in ( Figure 7A-B).

**********

-->6. PLOS authors have the option to publish the peer review history of their article (what does this mean? ). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy .-->

Reviewer #1: Yes: Sourbh Suren Garg

Reviewer #2: Yes: Saleem Iqbal

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures

You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation.

NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications.

Attachment

Submitted filename: PONE-D-25-34800_Comments.docx

pone.0343274.s002.docx (17.3KB, docx)
PLoS One. 2026 Feb 20;21(2):e0343274. doi: 10.1371/journal.pone.0343274.r002

Author response to Decision Letter 1

23 Dec 2025

RESPONSE TO REVIEWER COMMENTS

We thank the Referee for spending time and interest in our work and for helpful comments that will greatly improve the manuscript. We have checked all the general and specific comments provided by the Referee and have made all the necessary changes according to his instructions. Please refer to the yellow-highlighted sections or tracked changes in the revised manuscript.

Reviewer #1: The authors have conducted a good work addressing a relevant scientific problem and have presented their analyses clearly. A few constructive suggestions have been provided to enhance clarity and strengthen the overall quality of the manuscript.

1. The study uses very small sample sizes, which limits statistical confidence. The authors should acknowledge the limitation clearly in the Discussion and Conclusion.

Response: Thank you for your comment. We acknowledge the reviewer's point and conclude that our findings confidence in statistics is limited by the small sample sizes. In response, we have explicitly stated that findings should be read cautiously and verified in larger cohorts in both the Discussion and Conclusion sections.

2. The authors should strengthen the rationale for selecting PPARG as the primary hub gene. Ranking by degree score alone is not sufficient; additional biological justification and supporting literature linking PPARG specifically to PLOSL pathology should be provided.

Response: We appreciate the reviewer's emphasis on the necessity of bolstering the case for choosing PPARG as the main hub gene. We agree that ranking by degree alone is insufficient, even though PPARG was first found based on degree score. To address this, we included additional network analyses such as eccentricity, betweenness centrality, and closeness centrality. These analyses show that PPARG is not only highly connected but also plays a key role in regulating information flow and preserving network integrity within PLOSL-associated genes.

3. Survival analysis appears inappropriate because no PLOSL clinical survival dataset exists. The analysis should be removed or explicitly clarified as exploratory and non-disease-specific.

Response: We politely disagree with the reviewer statement. This survival study was conducted using the single-gene analysis module of the KMPlot web tool, which combines TCGA and GTEx expression data to evaluate gene-level associations. This analysis was carried out to assess the biological significance of hub genes, including PPARG, in more general neurodegenerative and systemic contexts, despite the lack of PLOSL-specific clinical survival information. In addition to confirming the functional consistency of these genes, this method offers initial insights that could direct future PLOSL pathology research.

4. The authors should reduce repetitive content, particularly in the Introduction and Discussion, to improve clarity and maintain a concise narrative.

Response: We appreciate the reviewer insightful suggestions. The introduction and discussion portions of the work have been meticulously updated to eliminate redundant material. To increase readability, preserve a succinct narrative, and improve clarity, redundant statements have been eliminated or combined.

5. The PCA and heatmap figures require improved resolution and clearer labeling to allow interpretability.

Response: Thank you for your suggestion. The PCA and Heatmap figures have been improved with high resolutions in the revised manuscript.

6. Figures should include clear color legends and larger, readable font sizes. Several elements, particularly those in white text, are difficult to interpret and should be improved for better visibility.

Response: We appreciate the reviewer insightful suggestions. All the figures have been improved with high resolutions in the revised manuscript.

7. The KEGG and GO enrichment results are listed but not fully interpreted. Please explain how identified pathways relate mechanistically to PLOSL progression, especially microglial and osteoclast dysfunction.

Response: We appreciate the reviewer insightful observation. We made it clear in the updated manuscript that PLOSL disease is directly associated with enhanced pathways like osteoclast differentiation and efferocytosis. While defective signaling in microglial promotes neuroinflammation and neurodegeneration, two major aspects of the disease, dysregulation of osteoclast differentiation contribute to the creation of bone cysts.

8. The text states that PPARG is associated with disease progression but does not cite primary mechanistic literature. Please add supportive references.

Response: Thank you for the valuable point. Some references regarding the claim made have been added to the revised manuscript discussion section.

9. The molecular docking results indicate moderate binding affinity, so the therapeutic relevance should be discussed cautiously. While the MD simulation plots (RMSD, RMSF, and Rg) are provided, the interpretation of these results should be expanded to clearly explain the structural stability and conformational behavior of the ligand–protein complexes.

Response: Thank you for your comment. The authors have added the therapeutic relevance with literature support in the molecular docking section for both the top drugs. The trajectories analysis of the MD simulation has been added in the revised manuscript with the structural stability and confirmational behavior of ligand-protein complexes as shown in Figure 10.

10. The chemical structures of AMG-131 and Elafibranor should be included as figures to support clarity.

Response: Thank you for the suggestion. The authors have added the chemical structure of both the compound in the revised manuscript as shown in figure 7.

11. The authors should also acknowledge the discontinued clinical trial status of Elafibranor and the experimental development stage of AMG-131 to avoid implying that these compounds are clinically viable options at this stage.

Response: We appreciate the reviewer bringing out this crucial point. We have updated the manuscript to make it clear that AMG-131 is still an exploratory chemical with no proven therapeutic value, and Elafibranor is no longer in clinical development. We now stress that these compounds are not yet therapeutically feasible possibilities and that our results are merely computational predictions that need to be thoroughly validated by experiments and preclinical research before being taken into consideration for clinical use.

12. The manuscript uses long sentences that reduce readability. Grammar, sentence structure, and transitions require careful language editing.

Response: Thank you for the comment. The paper has been revised for English language, grammatical mistakes and inappropriate long sentences and improved considerably in light of reviewer comment.

13. The Conclusion section should be moderated; currently it implies therapeutic development readiness. Please emphasize that experimental validation is still required.

Response: We are grateful to the reviewer for their feedback. There is no longer any suggestion of therapeutic preparedness in the Conclusion section. The revised language makes it very evident that all results are based on computational predictions and that further experimental validation is necessary before any therapeutic interpretation can be made on the possible significance of AMG-131, Elafibranor, and PPARG.

14. The authors may consider briefly discussing alternative hub candidates identified (e.g., C1QA, AIF1) to balance interpretation and avoid appearing to pre-select PPARG.

Response: We appreciate the reviewer recommendation. We have included a brief discussion in the revised manuscript of alternate hub genes possibilities including C1QA and AIF1 in the amended text. Although PPARG is still the major emphasis because of its biological significance and network centrality, considering these other hub genes offers a more balanced explanation and recognizes their possible roles in PLOSL disease.

15. The study would benefit from a short limitations paragraph summarizing dataset size, lack of clinical validation, and computational-only nature.

Response: we appreciate the reviewer suggestion. A brief restrictions paragraph has been added to the discussion. The small sample size of the accessible PLOSL datasets, the lack of experimental or clinical validation, and the study's computational only design are all highlighted in the updated text. This amendment makes it clear that the results are preliminary and need to be verified by larger datasets and biological validation.

16. The Introduction provides general background on PLOSL but could be more concise. Several paragraphs repeat known disease characteristics. Reducing redundancy will improve focus.

Response: We agree with the reviewer’s comment. The introduction has been shortened by removing repeated descriptions of PLOSL features and consolidating overlapping background information. This updated revision improves clarity and maintains a more concise and focused narrative in the updated manuscript.

17. The Introduction should state a clear research hypothesis and specific study objectives at the end. Currently, the transition into the aim of the study is broad and lacks direct specificity.

Response: We agree with the reviewer suggestion. The end of the Introduction has been revised in the updated manuscript to include a clear research hypothesis and specific study objectives, providing a more direct and focused transition into the aims of the study.

18. Authors are advised to incorporate more recent literature (within the past 5 years) on PPARG signaling in neuroinflammation and osteoclast regulation to better support the rationale for selecting PPARG as the hub gene in this study.

Response: Thank you for the valuable comment. Some recent literature on PPARG in neuroinflammation and osteoclast regulation is added to the revised manuscript discussion section.

19. Authors are advised to justify the selection of the specific GEO datasets used in this study, particularly regarding sample relevance and availability. Additionally, the differential expression threshold should be clarified, as the stated FDR < 2 is not standard practice; please correct or explain the appropriate adjusted p-value cut-off for clarity and consistency.

Response: We appreciate the reviewer's feedback. Since GSE3624 and GSE25496 are the only publicly accessible GEO datasets with verified PLOSL patient samples and suitable controls, they were chosen as the most pertinent and easily accessible datasets for this uncommon illness. Additionally, the differential expression threshold has also been adjusted. The analysis employed the standard adjusted p-value (FDR) < 0.05, which is now explicitly mentioned in the updated manuscript for correctness and consistency; the previously reported FDR < 2 was an error.

20. The PPI network analysis describes the construction process, but network confidence score choices (0.4 or 0.75) vary in different parts of the text. Standardize and justify the confidence threshold.

Response: We appreciate the reviewer pointing out the discrepancy. Throughout the entire work, we have now standardized the PPI network analysis to a 0.75 high-confidence criterion. Because it offers more consistent and biologically significant protein–protein interactions, this threshold was chosen. The manuscript has been updated accordingly.

21. The CytoHubba ranking selection uses degree only. Consider including at least one additional centrality parameter (e.g., closeness or betweenness) to strengthen hub gene selection.

Response: We appreciate the reviewer's insightful recommendation. Along with degree score, we have included other centrality metrics in the updated manuscript, such as eccentricity,between Centrality, and closeness centrality. These complementing measurements support the selection of PPARG and other hub genes by confirming their important locations in the network.

22. The volcano plot and heatmap would benefit from labeling the key differentially expressed genes to enhance interpretability and allow readers to easily identify the most significant up- and down-regulated genes.

Response: Thank you for the suggestion the volcano plot has been updated by labeling the differentially expressed genes to enhance interpretability and allow readers to easily identify the most significant up- and down-regulated genes in the revised manuscript.

23. The GO/KEGG tables list terms effectively, but narrative interpretation is brief. Highlight the most biologically relevant pathways in the context of PLOSL pathogenesis.

Response: We appreciate the reviewer comment. We have expanded the narrative to highlight the most relevant GO/KEGG pathways linked to PLOSL, particularly those involving microglial activation, immune dysregulation, lipid metabolism, and impaired phagocytic signaling, as they closely align with known mechanisms of PLOSL pathogenesis. The manuscript has been updated accordingly.

24. The presentation of docking results compares binding scores, but no comparison is made with known strong PPARG ligands beyond Rosiglitazone. Including known ligand reference values would contextualize the strength of interactions.

Response: Thank you for the suggestion. The authors have revisited the docking results and have added the control ligand Rosiglitazone results in the revised manuscript, along with the interactions as shown in Figure 8C

25. The Results section occasionally shifts into speculative statements (e.g., therapeutic implications) that belong in Discussion. Please separate objective findings from interpretation.

Response: The results section has been revised and such statements that were not required have been removed. However, some sentences that the authors felt were important for the continuous interpretation of the readers were retained.

RESPONSE TO REVIEWER COMMENTS

We thank the Referee for spending time and interest in our work and for helpful comments that will greatly improve the manuscript. We have checked all the general and specific comments provided by the Referee and have made all the necessary changes according to his instructions. Please refer to the yellow-highlighted sections or tracked changes in the revised manuscript.

Reviewer #2: The manuscript by Bokhari et al. presents a comprehensive computational study aimed at identifying key hub genes, pathways, and potential drug repurposing candidates for the rare and severe Nasu-Hakola disease (PLOSL). The authors employ a systems biology workflow, including differential gene expression analysis, protein-protein interaction (PPI) network construction, survival analysis, and sophisticated computational chemistry techniques like molecular docking and molecular dynamics (MD) simulations. The study's main claims are the identification of PPARG and AIF1 as central hub genes in PLOSL pathology and the proposal of AMG-131 and Elafibranor as promising therapeutic agents targeting PPARG. The significance of this work lies in its attempt to address a major therapeutic gap for a devastating rare disease using a cost-effective, bioinformatics-driven drug repurposing approach. The application of a 200 ns MD simulation to validate docking results adds considerable strength to the stability claims of the proposed drug-target complexes.

Well for authors must note that PLOS One encourages submissions across the methodological spectrum, but for a study making specific claims about therapeutic potential and biomarker identification, some form of experimental corroboration is increasingly expected to justify publication.

Response: Thank you for the valuable comment. As the paper scope is pure computational and we as a team mainly work on bioinformatics-related projects, the paper was submitted to PLOS ONE purely because it accepts computational work. The paper's conclusion can be validated by experimental validation; however, currently, we don’t have experimental facilities to validate the computational conclusions of this paper. The paper findings will speed up the PPARG as drug target as it may provide experimentalists to test this protein in a specific among the large dataset. The experimental

Attachment

Submitted filename: Response Letter.docx

pone.0343274.s004.docx (24.3KB, docx)

Decision Letter 1

Md Zubbair Malik

3 Feb 2026

<p>System Biology and Network-Based Approach to Identify the Therapeutic Signatures and Potential Inhibitors Against Polycystic Lipomembranous Osteodysplasia with Sclerosing Leukoencephalopathy

PONE-D-25-34800R1

Dear Dr. Alaa Abdulaziz Eisa.

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager®  and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support .

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Academic Editor

PLOS One

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

-->Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.-->

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

-->2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. -->

Reviewer #1: Yes

Reviewer #2: Yes

**********

-->3. Has the statistical analysis been performed appropriately and rigorously? -->

Reviewer #1: Yes

Reviewer #2: Yes

**********

-->4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.-->

Reviewer #1: Yes

Reviewer #2: Yes

**********

-->5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.-->

Reviewer #1: Yes

Reviewer #2: Yes

**********

-->6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)-->

Reviewer #1: Authors have well addressed most of the reviewer comments; however, the discussion of PPARG remains general. The manuscript does not clearly establish a mechanistic link between PPARG signaling and PLOSL-specific pathology (e.g., microglial/macrophage dysfunction, lipid homeostasis, and chronic inflammation). This section should be strengthened with clearer disease-specific interpretation and supporting literature before the manuscript can be considered for publication.

Reviewer #2: The revised version is significantly clearer, more rigorous, and better framed as a computational systems biology contribution. The improvements to the figures, network analysis, and molecular simulations are particularly commendable.

Congratulations!

**********

-->7. PLOS authors have the option to publish the peer review history of their article (what does this mean? ). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy .-->

Reviewer #1: Yes: Sourbh Suren Garg

Reviewer #2: Yes: Dr Saleem Iqbal

**********

Acceptance letter

Md Zubbair Malik

PONE-D-25-34800R1

PLOS One

Dear Dr. Eisa,

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS One. Congratulations! Your manuscript is now being handed over to our production team.

At this stage, our production department will prepare your paper for publication. This includes ensuring the following:

* All references, tables, and figures are properly cited

* All relevant supporting information is included in the manuscript submission,

* There are no issues that prevent the paper from being properly typeset

You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps.

Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing.

If we can help with anything else, please email us at customercare@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Md. Zubbair Malik

Academic Editor

PLOS One

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 Fig. Prior to normalization, the box plot, density plot, and mean expression plot are shown in (A), (B), and (C); following normalization, the same plots are shown in (D), (E), and (F). Using the Network Analyst Tool, variance-stabilizing normalization successfully removed noise from the data. The next methods used to analyze the data were Principal Component Analysis and the Quality Control Examination.

    (DOCX)

    pone.0343274.s001.docx (1.3MB, docx)
    Attachment

    Submitted filename: PONE-D-25-34800_Comments.docx

    pone.0343274.s002.docx (17.3KB, docx)
    Attachment

    Submitted filename: Response Letter.docx

    pone.0343274.s004.docx (24.3KB, docx)

    Data Availability Statement

    All relevant data are within the manuscript and its Supporting Information files.

    Articles from PLOS One are provided here courtesy of PLOS

    RESOURCES