Identification of Hub of the Hub-Genes From Different Individual Studies for Early Diagnosis,Prognosis,and Therapies of Breast Cancer

Abstract

Breast cancer (BC) is a complex disease, which causes of high mortality rate in women. Early diagnosis and therapeutic improvements may reduce the mortality rate. There were more than 74 individual studies that have suggested BC-causing hub-genes (HubGs) in the literature. However, we observed that their HubG sets are not so consistent with each other. It may be happened due to the regional and environmental variations with the sample units. Therefore, it was required to explore hub of the HubG (hHubG) sets that might be more representative for early diagnosis and therapies of BC in different country regions and their environments. In this study, we selected top-ranked 10 HubGs (CCNB1, CDK1, TOP2A, CCNA2, ESR1, EGFR, JUN, ACTB, TP53, and CCND1) as the hHubG set by the protein-protein interaction network analysis based on all of 74 individual HubG sets. The hHubG set enrichment analysis detected some crucial biological processes, molecular functions, and pathways that are significantly associated with BC progressions. The expression analysis of hHubGs by box plots in different stages of BC progression and BC prediction models indicated that the proposed hHubGs can be considered as the early diagnostic and prognostic biomarkers. Finally, we suggested hHubGs-guided top-ranked 10 candidate drug molecules (SORAFENIB, AMG-900, CHEMBL1765740, ENTRECTINIB, MK-6592, YM201636, masitinib, GSK2126458, TG-02, and PAZOPANIB) by molecular docking analysis for the treatment against BC. We investigated the stability of top-ranked 3 drug-target complexes (SORAFENIB vs ESR1, AMG-900 vs TOP2A, and CHEMBL1765740 vs EGFR) by computing their binding free energies based on 100-ns molecular dynamic (MD) simulation based Molecular Mechanics Poisson-Boltzmann Surface Area (MM-PBSA) approach and found their stable performance. The literature review also supported our findings much more for BC compared with the results of individual studies. Therefore, the findings of this study may be useful resources for early diagnosis, prognosis, and therapies of BC.

Keywords

Breast cancer hub of the hub-genes (hHubGs)early diagnosis and prognosis regulatory factors drug repurposing molecular docking

Introduction

Cancer is a malignant tumor that is initiated by multiple changes at the genetic and epigenetic levels that progressively lead to abnormal cell division and cellular alteration.^1,2 Breast cancer (BC) is a major public health concern, since it is the leading cause of cancer-related deaths in women worldwide. Breast cancer was the second most common cancer diagnosed, and the number was approximately 2 million (11.6% of all cancers) in 2018.³ Breast cancer became the leading cause of diagnosis in 2020, and the number was approximately 2.2 million (11.7% of all cancers) worldwide.⁴ Thus, it is a matter of concern that the number of patients with BC and deaths is increasing every year, despite claims that there have been significant improvements in BC treatment. According to the American Cancer Society, the 5-year survival rates of patients with BC are 98%, 93%, 72%, and 22% in stage 1, stage 2, stage 3, and stage 4, respectively.⁵ The 5-year BC survival rate is the percentage of people who survive at least 5 years after diagnosis of BC. From a previous study, it was found that around 52.34% of the patients with BC received treatment in stage 4, 32.81% in stage 3, 10.16% in stage 2, and only 4.69% patients were treated in stage 1.⁶ It has been observed that highest number of patients is detected in the fourth stage of BC and their survival rate is the lowest. Therefore, early diagnosis and therapeutic improvement is essential to reduce the death rate of patients with BC. The BC-causing genes may play the key role in this regard.

There are several individual studies on transcriptomics (microarray/RNA-Seq) analysis in the literature that suggested BC-causing hub-genes (HubGs) to investigate the pathogenetic processes of BC. We systematically reviewed 74 individual articles from different online sources that suggested BC-causing HubGs and found that their HubG sets are not so similar with each other (see Supplemental Table S1). It may be happened due to the regional and environmental variations with the experimental units. We found a total of 259 different HubGs from those articles in which there was no any common HubGs among the articles. So, it may be difficult to take a common treatment plan for all patients with BC in different regions and environment. Therefore, it is required to explore hub of the HubG (hHubG) sets that may be more representative BC-causing HubGs for diagnosis and therapeutic improvement for the treatment against BC in different environments. Exploring BC-causing hHubGs out of 259 HubGs highlighting their early diagnostic, prognostic, and therapeutic characteristics by the wet-laboratory experimental procedure may be laborious, time-consuming, and costly. In this study, we attempted to explore BC-causing hHubGs highlighting their early diagnostic, prognostic, and therapeutic characteristics by using the integrated statistics and bioinformatics approaches.

To explore disease-causing HubGs, the protein-protein interaction (PPI) network analysis is a widely used popular approach.^7-12 In this study, we also explored BC-causing hHubGs by the PPI network analysis of 259 HubGs that we mentioned earlier. In the case of drug discovery, drug repurposing (DR) is a promising strategy, where new therapeutic inventions are made through existing drugs.¹³ It is secure, more efficient, less expensive, and less time-consuming compared with the de novo process due to it skip several steps from the target-based drug selection to clinical validation.¹⁴ Therefore, in this study, we detected hHubGs-guided repurposable drug molecules. There are several steps involved in the in silico analysis of the DR process, such as network pharmacology, molecular docking analysis, and dynamic simulation analysis.¹⁵ Network pharmacology is an efficient tool for systematic pharmacologic research which creates interactions between target genes and drugs through network-based methods.^16,17 Molecular docking analysis plays a vital role in drug design through structural binding between therapeutic targets and drug molecules by calculating affinity scores.¹⁸ The molecular dynamic (MD) simulation indicates the stability of docking patterns. Recently, it has been widely used and has become one of the most popular platforms to validate the candidate drugs.^19,20 The workflow of this study is displayed in Figure 1. It should be mentioned here that the similar workflow was also considered in some previous studies for other diseases.^21,22

Figure 1.

The workflow of this study.

Methods and Materials

Data collection on BC-causing HubG sets

We collected HubG sets and drug molecules as molecular metadata from different published articles up to December 31, 2023, through online search strategy with the keywords “breast cancer [ti], gene expression, RNA-Seq, scRNA-Seq, differentially expressed genes (DEGs), and core/key/hub genes” from the public sources including PubMed in the National Center of Biotechnology Information (NCBI) (https://pubmed.ncbi.nlm.nih.gov/), Google Scholar (https://scholar.google.com/), and Google (https://www.google.com/). We found a total of 74 independent articles that suggested BC-related HubGs-set. We collected 74 HubG sets from the selected 74 articles. We found a total of 259 unique HubGs from those HubG sets and used them to explore hHubGs for further investigation in this study (see Supplemental Table S1).

Data collection on drug molecules

We collected candidate drug molecules from 3 platforms: (1) published articles, (2) the Gene Set Cancer Analysis (GSCALite) database,²³ and (3) the Drug Gene Interaction Database (DGIdb).²⁴ Some of our selected articles suggested candidate drugs for the treatment of BC and considered them set A.^25,26 We collected meta-drugs from 2 databases DGIdb and GSCALite through the hHubG-drugs interaction and considered them sets B and C, respectively (Supplemental Table S2).

Detection of BC-causing hub of the HubGs

Network-based strategies are widely used in identifying disease-driving KGs and their potential regulators.^27,28 We submitted all unique HubGs to the STRING web tool (Search Tool for the Retrieval of Interacting Genes) (version: 11.5) to construct PPI network.²⁹ Protein-protein interactions of the all HubGs were constructed with a confidence score 0.90. Subsequently, Cytoscape software (version: 3.9.0) was used for visualization of the PPI network.³⁰ The “Analyze Network tool” in Cytoscape was used for topological degree analysis, and the degree >90 was considered as the cutoff criterion for hHubG identification.

Assessment of hHubGs as the early diagnostic and prognostic biomarkers

To assess hHubGs as early diagnostic and prognostic biomarkers, we performed their expression analysis with the different stages (normal status, stage 1, stage 2, stage 3, and stage 4) of BC progression by using box plots through the UALCAN web-tool and The Cancer Genome Atlas (TCGA) database.³¹ To investigate the prognostic power of hHubGs, we also developed BC prediction models through 2 popular machine learning methods known as support vector machine (SVM) and random forest (RF). To perform prediction models, we collected 3 microarray gene-expression data sets with accession numbers GSE65216,³² GSE10810,³³ and GSE36295³⁴ from the Gene Expression Omnibus (GEO) database. Each data set divided into 2 groups: training data (70% of all data) and test data (30% of all data). Finally, the receiver operating characteristic (ROC) curve was used to assess the prognosis performance. The ROC curve was constructed using the R package ROCR. We also used independent database “DisGeNET” and web-tool “Enrichr” to verify the association of hHubGs with BC and other disease by disease-hHubGs enrichment analysis.³⁵

Functional and pathway enrichment analysis

To investigate the pathogenetic processes of hHubGs, we performed their Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genome (KEGG) pathway enrichment analysis. The GO enrichment analysis was performed in 3 categories namely biological process (BP), cellular component (CC), and molecular function (MF). Gene Ontology functional and KEGG pathway enrichment analyses were performed using the online database DAVID (version: 6.8) online tool.³⁶

Regulatory interaction network

We performed regulatory interaction network to identify the key transcription factors (TFs) and microRNAs (miRNAs) as the transcriptional and posttranscriptional regulators of hHubGs. The online tool “NetworkAnalyst” (version: 3.0) was used to construct the regulatory interaction network.³⁷ ChEA and TarBase (v8.0) databases were selected for constructing hHubG-TF and hHubG-miRNA interaction networks, respectively.

Drug repurposing

We used molecular docking analysis, a popular in silico validation technique, to perform the binding between the drug and target. We considered hHubGs as targets (receptors/proteins) and meta-drug agents collected from published articles and the online databases GSCALite and DGIdb as drugs (compounds) for performing molecular docking analysis. We collected 3-dimensional (3D) structures of proteins from the Protein Data Bank (PDB) (https://www.rcsb.org/)³⁸ and the SWISS-MODEL (https://swissmodel.expasy.org/)³⁹ databases due to the required to perform molecular docking analysis. The 3D structures of all compounds were downloaded from the PubChem (https://pubchem.ncbi.nlm.nih.gov/) database.⁴⁰ PyMOL (PyMOL2) software was used for the preprocessing of 3D structure proteins.⁴¹ PyRx software was used to perform the blind molecular docking analysis between proteins and compounds and selected shortlisted drugs based on binding affinity scores (BASs) (kcal/mol).^42,43 Then, we validate the selected drug molecules with the active sites (by using PyMOL2 software)⁴¹ of our selected hHubGs through flexible docking analysis. The 3D interaction between the protein and compound was constructed and visualized by using USCF Chimera and Discovery Studio Visualizer 2021 software.⁴⁴

MD simulation studies

Molecular dynamic simulations were performed using YASARA Dynamics software⁴⁵ and the AMBER (Assisted Model Building with Energy Refinement) force field⁴⁶ to test the dynamic behavior of the top protein-compound interactions. The AMBER’s force fields have been extensively validated and optimized for various biological molecules, making it a robust choice for simulation study.⁴⁷ The root mean square deviation (RMSD) value between the crystal reference and the predicted structure is commonly used to confirm whether a close-match docked pose was predicted or not by docking simulations. If the RMSD value is ⩽2 Å, the result is considered as good result.⁴⁸ Three top-ranked interactions based on affinity scores were selected to perform MD simulations. For simulation performance, the protein-compound interaction hydrogen bonding network is optimized and solvated through the TIP3P water model in a simulation cell.⁴⁹ The periodic boundary condition was maintained by a solvent concentration of 0.997 g L⁻¹. pKa was calculated during solvation by subjecting to titratable amino acids in the interaction. A time-step interval of 2.50 fs (298 K, pH = 7.4, 0.9% NaCl) under physiological conditions was used to run several time-step algorithms for each simulation.⁵⁰ The SETTLE and LINCS algorithms were used to contain water molecules and constrain all bond lengths, respectively.^51,52 For further analysis, trajectories were recorded every 250 ps, and subsequent analyses were implemented by default scripts of YASARA Macro and SciDAVis software.⁵³ Then, all snapshots were subjected to the Molecular Mechanics Poisson-Boltzmann surface area (MM-PBSA) in YASARA software to calculate the binding free energy using the given equation.⁵⁴

B i n d i n g f r e e E n e r g y Δ G = G_{b o u n d} - G_{u n b o u n d}

Here, YASARA built-in macros were used with consideration AMBER 14 as a force field to calculate the MM-PBSA binding energy, where a greater positive energy indicates better binding.

Results

Exploring BC-causing hub-genes by the systematic literature review

We found a total of 297 BC-related articles from various online sources (PubMed, Google Scholar, and Google) by keywords search up to December 31, 2023. Finally, we selected 74 articles out of them through the inclusion-exclusion criteria (see Supplemental Figure S1 and Supplemental Table S1). Then, we detected a total of BC-causing 259 different HubGs from those 74 HubG sets as described in the “Materials and Methods” section and used them for further investigation in this study.

Detection of hub of the HubGs

The PPI network was constructed based on those 259 individual HubGs to detect BC-causing hHubGs as key genomic biomarkers. It included 209 of nodes, 5785 edges, average node degree 39.5, average local clustering coefficient 0.599, and P < 1.0e-15. The PPI network was visualized in Figure 2, where hHubGs are indicated by large-size nodes and selected based on the top degree of connectivity (degree >95). We selected top-ranked 10 hHubGs (CCNB1, CDK1, TOP2A, CCNA2, ESR1, EGFR, JUN, ACTB, TP53, and CCND1) in which 6 hHubGs (CCNB1, CDK1, TOP2A, CCNA2, ESR1, and EGFR) belong to the set of already published BC-causing top-ranked 10 HubGs (see Table 1 and Supplemental Table S1).

Figure 2.

The PPI network based on 259 unique HubGs, where large size indicates hHubGs with the highest degree of connectivity (degree >90).

Table 1.

List of top-ranked 10 BC-causing HubGs/hHubGs based on the agreement of published articles (Supplemental Table S1) (A*) and hHubGs through PPI network analysis (B*) (Figure 2), respectively.

Top-ranked 10 HubGs by the agreement of published articles (Supplemental Table S1)		Top-ranked 10 hHubGs (proposed)		Common hHubGs $(A^{} \cap B^{})$
Genes (A*)	Number of agreed articles	Genes (B*)	Degree of PPI score	Common hHubGs $(A^{} \cap B^{})$
TOP2A	8	TP53	143	CCNB1 CDK1 TOP2A CCNA2 ESR1 EGFR
CDK1	7	ACTB	122
CCNB1	9	EGFR	102
FN1	8	CCND1	100
RRM2	6	JUN	100
BUB1B	7	ESR1	95
CCNB2	8	CDK1	99
CCNA2	8	CCNB1	99
ESR1	6	CCNA2	98
EGFR	6	TOP2A	91

Bold indicates common genes.

Performance of hHubGs as early diagnostic and prognostic biomarkers

To investigate the significance of differential expression patterns of top-ranked 10 hHubGs between the normal group and 4-stage group (stage 1, stage 2, stage 3, and stage 4) of BC individually, we performed box-plot analysis using t test statistic with P < .001 based on their TCGA expression profiles (Figure 3A). We observed that all hHubGs individually showed significantly differential expression patterns (P values ranged from 8.99E-04 [TP53] to 2.22E-16 [ACTB]) between stage 1 BC and control samples group. Thus, our identified hHubGs may play an important role in the early stage diagnosis of BC. To investigate the prognostic power of top-ranked 10 hHubGs, we developed RF and SVM-based 2 BC prediction models by using the gene-expression data with NCBI-GEO accession number GSE65216 as training data set, and GSE10810 and GSE36295 as the test data sets. To investigate the performance of the prediction models, we constructed ROC curves in Figure 3B, where green indicates training performance and red and blue indicate test performance. We observed that both the training and test performances were good with area under the curve (AUC) values >0.86 for both the RF and SVM-based prediction models. Furthermore, we observed that hHubGs are significantly associated with invasive carcinoma of BC and luminal B breast carcinoma (Supplemental Figure S2). Thus, our suggested hHubGs are significantly associated with almost all subtypes of BC.

Figure 3.

(A) Analysis of expression levels of hHubGs between the normal group and 4-stage group (stage 1, stage 2, stage 3, and stage 4) of BC, and (B) ROC curves to investigate the prognostic power of hHubGs by the RF and SVN-based BC prediction models with 3 gene-expression profile data sets in NCBI-GEO database.

HubGs-set enrichment analysis

The GO functional and KEGG enrichment analysis showed that unique HubGs were enriched by 234 GO-BP terms, 56 GO-MF terms, 38 GO-CC terms, and 85 pathways. Then, we selected the significant GO terms and KEGG pathways that were associated with at least 4 of our identified hHubGs and P < .01. Four BP terms (positive regulation of transcription from RNA polymerase II promoter, cell proliferation, cell division, and regulation of cell cycle), 5 CC terms (nucleoplasm, nucleus, cytosol, cytoplasm, and membrane), 5 MF terms (protein binding, enzyme binding, adenosine triphosphate (ATP) binding, TF binding, and protein heterodimerization activity), and 7 KEGG pathways (pathways in cancer, PI3K-Akt signaling pathway, cell cycle, herpes simplex infection, human T-lymphotropic virus type 1 (HTLV-I) infection, proteoglycans in cancer, and PI3K-Akt signaling pathway) were associated with at least 4 hHubGs and may be responsible for BC progression in Table 2.

Table 2.

Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways enrichment analysis results of all HubGs highlighting number of HubGs (nHubGs), P value, and associated hub of the HubGs (hHubGs).

ID	Terms/pathways	HubGs	P	Associated hHubGs
GO: Biological process (BP)
GO:0045944	Positive regulation of transcription from RNA polymerase II promoter	66	3.8E−21	TOP2A, TP53, EGFR, ESR1, JUN
GO:0008283	Cell proliferation	38	5.0E−18	EGFR, CDK1, TP53, JUN
GO:0051301	Cell division	36	5.9E−17	CCNB1, CCND1, CCNA2, CDK1
GO:0051726	Regulation of cell cycle	23	2.8E−16	JUN, CCNB1, TP53, ESR1
GO: Cellular components (CCs)
GO:0005654	Nucleoplasm	123	2.6E−26	ACTB, CCND1, TP53, CCNB1, ESR1, JUN, CDK1, TOP2A, CCNA2
GO:0005634	Nucleus	175	5.9E−24	CCND1, TP53, CCNB1, JUN, ESR1, CDK1, TOP2A, CCNA2, EGFR
GO:0005829	Cytosol	129	1.0E−22	ACTB, CCND1, TP53, CCNB1, JUN, CDK1
GO:0005737	Cytoplasm	154	8.7E−16	ACTB, CCND1, TP53, CCNB1, ESR1, CDK1, TOP2A, CCNA2, EGFR
GO:0016020	Membrane	73	1.1E−08	ACTB, CCND1, EGFR, CCNB1, ESR1, CDK1
GO: Molecular functions (MFs)
GO:0005515	Protein binding	245	2.6E−29	ACTB, TP53, JUN, TOP2A, EGFR, CCND1, CCNB1, ESR1, CDK1, CCNA2
GO:0019899	Enzyme binding	33	3.9E−15	TOP2A, EGFR, CCND1, JUN, ESR1, TP53
GO:0005524	ATP binding	67	9.8E−13	TOP2A, ACTB, TP53, EGFR, CDK1
GO:0008134	Transcription factor binding	24	1.0E−09	JUN, ESR1, CCND1, TP53
GO:0046982	Protein heterodimerization activity	28	4.8E−08	TOP2A, EGFR, JUN, TP53
KEGG pathways
hsa05200	Pathways in cancer	51	1.4E−20	CCND1, TP53, EGFR, JUN
hsa04151	PI3K-Akt signaling pathway	33	1.6E−20	CCND1, JUN, CCNA2, TP53
hsa04110	Cell cycle	29	2.6E−18	CCNB1, CCND1, CCNA2, CDK1, TP53
hsa05168	Herpes simplex infection	24	1.0E−16	EGFR, CCND1, TP53, JUN
hsa05166	HTLV-I infection	36	1.8E−15	CCND1, JUN, TP53, ESR1
hsa05205	Proteoglycans in cancer	31	2.4E−14	EGFR, ACTB, CCND1, ESR1, TP53
hsa04151	PI3K-Akt signaling pathway	40	3.1E−14	EGFR, CCND1, TP53, TOP2A

Regulatory network analysis

To identify the top-ranked transcriptional and posttranscriptional regulators of hHubGs, we constructed and visualized KG-TF and KG-miRNA interaction networks in Figure 4A and B, where pink indicates hHubGs and green indicates TFs and miRNAs in Figure 4A and B, respectively. We observed that 5 TFs (MYC, HNF4A, KLF4, POU5F1, and SOX2) were associated with at least 8 hHubGs (degree ⩾8), so these 5 TFs considered as transcriptional factors of hHubG. Similarly, 5 miRNAs (hsa-mir-103a-3p, hsa-mir-107, hsa-mir-16-5p, hsa-mir-34a-5p, and hsa-mir-23b-3p) were associated with at least 8 hHubGs (degree ⩾8), so these 5 miRNAs were considered as posttranscriptional factors of hHubGs. We found from EMBL-EBI (European Molecular Biology Laboratory—European Bioinformatics Institute) database that 3 hHubGs (JUN, TP53, and ESR1) out of 10 hHubGs are TFs and indicated them by diamond shape in Figure 4. Figure 4C(I) indicates that 3 TFs of hHubGs (MYC, KLF4, and POU5F1) showed downregulation, and 2 TFs (HNF4A and SOX2) showed upregulation. Figure 4C(II) shows that KLF4 is a upstream and SOX2 is a downstream TFs for the JUN hHubGs (TFs), so JUN acts as an inhibitor to both KLF4 and SOX2 (if the expression level of the downstream TF decreased, then core TF acted as an activator otherwise inhibitor). Similarly, TP53 acted as an activator to POU5F1 and inhibitor to MYC and SOX2, and ESR1 acted as an activator to KLF4.

Figure 4.

hHubGs-regulatory networks, where pink indicates hHubGs (diamond shape hHubGs indicate TFs themselves): (A) hHubGs-TFs interaction network, where large green indicates key transcription factors; (B) hHubG-miRNA interaction network, where large green indicates key posttranscriptional factors; and (C) upstream and downstream analysis of TFs (I) differential expressions for TFs of hHubGs (II) upstream and downstream regulation by hHubGs that are also TFs (JUN, TP53, and ESR1), where pink diamond indicates core TFs, red rectangle indicates upstream TFs, and blue rectangle indicates downstream TFs.

Drug repurposing

First, we collected BC and hHubGs associated with a total of 258 drug agents from the published articles and 2 online databases: DGIdb and GSCALite. Then, we performed blind molecular docking analysis (entire protein) between compounds (drug agents) and target proteins (hHubGs). We selected the top-10 poses based on the smallest docking scores between compounds and target proteins. To check the quality of binding poses, we calculated the RMSD between native pose and our sleeted 10 poses. Finally, we selected the best pose based on the smallest RMSD and considered the docking score in that pose. We visualized the molecular docking results of the top 125 interactions in Figure 5A, where the y-axis represents the target proteins and x-axis compounds and different colors indicate the affinity scores. Compounds were arranged in x-axis in ascending order based on the average BAS across all proteins. For example, top-ranked 10 compounds (SORAFENIB, AMG-900, CHEMBL1765740, ENTRECTINIB, MK-6592, YM201636, masitinib, GSK2126458, TG-02, and PAZOPANIB) bind to the receptor proteins with average BAS less than −8.76 kcal/mol (binds strongly to the receptor proteins), and the next top-ranked 10 compounds (GSK-269962A, MLN-8054, ENMD-2076, ENMD-981693, NERVIANO, SB590885, Lapatinib, GSK1070916, TAK-715, and PF-03814735) bind to the receptor proteins with average BAS greater than −8.23 kcal/mol (binds weakly to the receptor proteins). Therefore, top-ranked 10 compounds (SORAFENIB, AMG-900, CHEMBL1765740, ENTRECTINIB, MK-6592, YM201636, masitinib, GSK2126458, TG-02, and PAZOPANIB) were proposed as candidate drugs against BC. In Figure 5B, the BAS of top-ranked 10 compounds with all receptors was highlighted. We found 6 overlapping genes between the top-ranked HubGs and hHubGs in Figure 5C. We cross-validated the proposed candidate drugs with 4 uncommon top-ranked HubGs in Figure 5D. To validate the top-ranked drug molecules with active sites of top-ranked HubGs by their flexible docking analysis, we predicted their active sites by the PyMOL software and listed in Table S3. Then, we perform flexible docking analysis between drugs and selected top-ranked HubGs (with flexibility of active sites) in Table 3. We also found a strong bond between them, so we may recommend our candidate drugs for universal use against BC.

Figure 5.

Molecular docking analysis results for screening candidate drugs against BC: (A) presented interaction binding affinity scores between hHubGs and drug agents; (B) visualization of interaction between 10 candidate drugs and hHubGs by zooming; (C) overlap between the top-ranked HubGs and hHubGs; and (D) cross-validation of the candidate drugs by top-ranked HubGs.

Table 3.

Flexible docking scores between the proposed drug molecules and the active sites of hHubGs.

Proteins	AMG900	CHEMBL1765740	SORAFENIB	ENTRECTINIB	MK6592	YM201636	masitinib	GSK2126458	TG02	PAZOPANIB
TOP2A	−8.2	−7.3	−7.1	−7	−7.5	−7.3	−6.8	−7.1	−6.9	−6.3
EGFR	−9.3	−8.5	−7.3	−7.6	−7.8	−8	−7.1	−6.9	−7.1	−6.1
ESR1	−8.9	−9.3	−9.5	−8.3	−8	−6.6	−7.3	−7.2	−7	−6.6
TP53	−8.5	−8.8	−7.8	−8	−8.1	−7.9	−7.6	−6.7	−7.1	−7.8
CDK1	−7.7	−6.8	−7.5	−7.6	−7.1	−7.2	−7	−6.7	−6.5	−6.8
CCND1	−7.6	−7.4	−8.1	−7.7	−7.3	−7.4	−7.6	−7.7	−7.4	−7.1
JUN	−8.5	−7.7	−8.1	−8.1	−7.7	−7.1	−6.7	−7	−6.9	−6.7
ACTB	−9.2	−7.8	−8.6	−7.2	−7.5	−7.7	−7.1	−7.7	−7.1	−6.6
CCNB1	−8.3	−7.5	−7.5	−8.1	−7.7	−7.6	−7.4	−7	−6.7	−7.2
CCNA2	−9.6	−7.6	−8.5	−7.5	−7.6	−7.2	−6.9	−7.5	−5.8	−6.3

Molecular dynamic simulations

Three top interactions, SORAFENIB-ESR1 (BAS = −12.5), CHEMBL1765740-EGFR (BAS = −10.8), and AMG900-TOP2A (BAS = −12.4), between candidate drugs and target proteins were selected for stability analysis using 100-ns MD-based MM-PBSA simulations. We noticed that all systems were remarkably stable in variations of moving and initial drug-target interactions in Figure 6. We calculated the RMSD of selected interactions to check the stability of structure during the simulation period in Figure 6A. All the systems projected RMSD range from 1 Å to 1.75 Å, .75 Å to 2.25 Å, and .5 Å to 2.75 Å for SORAFENIB-ESR1, CHEMBL1765740-EGFR, and AMG900-TOP2A interactions, respectively. We observed that all interactions fluctuated slightly between 0 and 16 000 ps and were stable in the remaining simulations. We also calculated the MM-PBSA binding energy for the 3 selected top interactions in Figure 6B. The average binding forces for SORAFENIB-ESR1, CHEMBL1765740-EGFR, and AMG900-TOP2A interactions were 217.5, 130.45, and 131.4 kJ/mol, respectively.

Figure 6.

MD simulation analysis of the top 3 interactions, where blue, red, and green indicate SORAFENIB-ESR1, CHEMBL1765740-EGFR, and AMG900-TOP2A interactions, respectively: (A) time evolution of RMSDs of backbone atoms (C, Cα, and N) for protein for each docked complex, and (B) represented binding free energy of each snapshot for the top 3 interactions.

Discussion

This study analyzed metadata on BC-causing HubGs collected from different published articles. At first, we selected 74 articles out of 297 by the systematic review as shown in Supplemental Figure S1. Different 259 HubGs were found in total from the selected articles (Supplemental Table S1), and these data were used for further analysis in this study. Top-ranked 10 HubGs (CCNB1, CDK1, TOP2A, CCNA2, ESR1, EGFR, JUN, ACTB, TP53, and CCND1) were selected as the hHubGs out of different 259 HubGs by the PPI network analysis, where 6 hHubGs (CCNB1, CDK1, TOP2A, CCNA2, ESR1, and EGFR) are common with the top-ranked HubGs based on the agreement of published articles (Table 1). Then, we verified their differential expression patterns between BC and control samples using independent data sets from NCBI and TCGA databases to the prediction models and box-plot analysis. We observed that the proposed hHubGs can be used as the strong biomarkers for early diagnosis and prognosis, where 8 genes (CCNB1, CDK1, TOP2A, CCNA2, ESR1, ACTB, TP53, and CCND1) are up-regulated hHubGs and the rest 2 genes (EGFR and JUN) are down-regulated hHubGs. In particularly, the CCNB1 gene is associated with aggressive tumor behavior, larger tumor size, higher grade, hormone receptor negativity, HER2 positivity, and poor clinical outcome.⁵⁵ The TOP2A gene has been suggested in 8 articles that this gene plays a prognostic role in patients with BC.^56-62 Overexpression of TOP2A is a marker of poor prognosis, especially in luminal/hormone receptor-positive BC, and may serve as both a prognostic and predictive biomarkers in BC disease.⁶³ Seven articles claim that the CDK1 gene is responsible for the occurrence and development of BC.^{26,57-59,64,65} CDK1 is associated with shorter overall survival, progression-free survival, and relapse-free survival, as well as more advanced clinical stage of cancer.⁶⁶ The CCNA2 gene has been identified in 6 studies as a promising biomarker in silico analysis related to BC progression.^{26,56,57,67-72} The CCNA2 gene is linked to a poor prognosis and resistance to endocrine therapy and serve as a prognostic biomarker in estrogen receptor-positive patients with BC.⁷³ The ESR1 gene suggested in several articles suggested as generalized BC.^56,74-78 The EGFR gene and its signaling pathways are related to the progression of BC and can be used as a therapeutic target against BC therapy.^61,74,79-81 Overexpression of EGFR is associated with more aggressive, ER-negative disease and poorer prognosis, particularly in triple negative BC.⁸² The JUN gene has been suggested that it can be used as a biomarker for the diagnosis and therapeutic target of BC.^74,83,84 In addition, 3 genes (ACTB, TP53, and CCND1) are supported by 3 articles.^{25,74,75,77,79,85-89} For more details, please see Figure 7. The CCND1 gene encodes cyclin D1, with a critical role in BC prognosis.⁹⁰ We observed that our proposed HubGs are mostly associated with a poor prognosis in patients with BC and also suggested for generalized BC by almost all of the 74 articles we listed in Supplemental Figure S1.

Figure 7.

Agreement of the proposed BC-related hHubGs with other independent studies.

To investigate the pathogenetic processes of hHubGs, we selected the top significant terms for each BP, MF, CC, and KEGG pathway associated with at least 4 hHubGs and P < .01 in Table 2. The GO terms and KEGG pathways related to our identified hHubGs have suggested several studies that are directly linked to BC progression. In particularly, the positive regulation of transcription from RNA polymerase II promoter term with BC cells associates late events in transcription with repair processes in eukaryotic cells.⁹¹ The regulation of cell cycle and cell proliferation are important indicators of BC prognosis.⁹² The cell proliferation employs transcriptional and nontranscriptional mechanisms to activate the cascade of cyclin-dependent kinases in BC.⁹³ The asymmetric cell division intricately regulates the diverse states of stem cells, pivotal in orchestrating the initiation of BC.⁹⁴ Observing the spatial overlap between BRCA1 and nucleus presents intriguing implications for the interplay between BC proteins within the nucleoplasm-nucleolus pathway, shedding light on their potential functional significance.⁹⁵ Several proteins are detected in both cytosolic and membrane fractions, exhibiting robust associations with distinct subfamilies of BC.⁹⁶ The TF binding term may lead to new investigations to understand the mechanism of CXXC5 in BC and may suggest new treatments for ER + BC.⁹⁷ The protein-binding term associated with BC-causing DEGs.⁷⁸ Two terms, enzyme binding and ATP binding, proposed by other studies that are important for BC.^98,99 The KEGG term PI3K-Akt signaling pathway proposed for BC cancer by bioinformatics study.¹² Furthermore, 3 KEGG pathways are cell cycle, HTLV-I infection, and proteoglycans in cancer are also related to BC progression claimed by other studies.^100-102

Finally, we suggested hHubGs-guided 10 candidate drugs molecules (SORAFENIB, AMG-900, CHEMBL1765740, ENTRECTINIB, MK-6592, masitinib, GSK2126458, TG-02, YM201636, and PAZOPANIB) for the treatment of patients with BC. These drug molecules might be effective for generalized patients with BC in different regions and environments, since these molecules strongly bind to all target genes/proteins (hHUBGs). It should remind here that hHubGs are the representative of all HubGs that were detected in different regions and environment (column 3 in Supplemental Table S1). However, among the identified candidate drugs, some drugs are already approved by the Food and Drug Administration (FDA) for the treatment of other diseases, some are under clinical trials and almost all drugs are supported by other in silico studies for BC. The mechanism of action of our proposed drug is also briefly described below. In particular, SORAFENIB was approved by the FDA in 2013 for the treatment of thyroid carcinoma¹⁰³ and also is under phase 2 clinical trials for the treatment of stage 4 BC.^104,105 The drug SORAFENIB is a multikinase inhibitor that provides antiproliferative, antiangiogenic, and antimetastatic effects by targeting cell surface tyrosine kinase receptors and downstream intracellular kinases implicated in tumor cell proliferation.¹⁰⁶ The drug AMG-900 is not yet approved for the treatment of any disease but it is under clinical trial (phase 1) for advanced solid tumors.¹⁰⁷ AMG-900 is an aurora kinases inhibitor, an essential regulator of cell division in mammalian cells, Aurora-A and Aurora-B expression and kinase activity is elevated in a variety of human cancers and is associated with high level of proliferation and poor prognosis.¹⁰⁸ The drug CHEMBL1765740 has not yet been published or is not under clinical trial for any disease but our study shows good efficacy in both molecular docking and MD simulation analysis. The drug ENTRECTINIB for the treatment of BC by Japan in 2019.¹⁰⁹ Entrectinib is a tyrosine kinase inhibitor that has shown potent antitumor effects in preclinical studies and has been found to inhibit tumor growth and survival in a dose-dependent manner in murine models.¹¹⁰ The drug masitinib was approved by the European Union (EU) for the treatment of amyotrophic lateral sclerosis and mast cell disease and was later denied approval in 2017 and 2018.^111,112 The drug GSK2126458 is under clinical trial of phase 1 study in human patients with BC.¹¹³ The primary mechanism of action of GSK2126458 is inhibition of the PI3K-Akt signaling pathway, which is critical for cancer cell proliferation and survival.¹¹⁴ The drug TG-02 has been approved by the FDA as an orphan drug for glioma and has also been suggested as an enzyme inhibitor of CDK for the treatment of cancer disease.^115,116 The TG-02 and YM201636 drugs have been also suggested for treatment against BC through molecular docking analysis.¹² TG-02 reduces cell proliferation and promotes cell death in cancer by inhibiting multiple kinases, particularly CDK1, CDK2, CDK7, and CDK9.¹¹⁷ YM201636 inhibits the phosphatidylinositol 3,5-bisphosphate kinase (PIKfyve), and has been shown to have antiproliferative effects on liver cancer cells in a dose-dependent manner.¹¹⁸ The drug PAZOPANIB is under clinical trial (phase 2) for BC treatment.¹¹⁹ Pazopanib is an oncostatic drug with antiangiogenic properties via inhibition of the intracellular tyrosine kinase vascular endothelial growth factor receptors (VEGFR).¹²⁰ We conclude from the literature review that the drug ENTRECTINIB have been approved and three drugs (SORAFENIB, GSK2126458, and PAZOPANIB) are under the clinical trial for BC treatment and other 6 drugs (AMG-900, CHEMBL1765740, MK-6592, masitinib, TG-02, and YM201636) are approved/under clinical trials for different diseases. Thus, we strongly recommend for further evaluation at the molecular level by experimental-laboratory testing and hope that effective results can be obtained.

Conclusions

This study identified most representative top-ranked 10 BC-causing HubGs (CCNB1, CDK1, TOP2A, CCNA2, ESR1, EGFR, JUN, ACTB, TP53, and CCND1) as the hHubGs by the PPI network analysis of 259 individual HubGs that are obtained from 74 previously published individual studies. Breast cancer prediction analysis based 3 gene-expression profiles (GSE65216, GSE10810 and GSE36295) of NCBI and the box-plot analysis with the gene-expression profiles of TCGA databases confirmed the differential expression patterns of hHubGs between BC and control samples. The enrichment analysis with GO terms and KEGG pathways revealed some crucial BC-causing BP (regulation of cell cycle and cell proliferation), MFs (nucleoplasm and nucleus), CCs (protein binding and enzyme binding), and pathways (PI3K-Akt signaling pathway and cell cycle) that we mentioned in the “Discussion” section. Gene regulatory network (GRN) disclosed top-ranked 5 TFs (MYC, HNF4A, KLF4, POU5F1, and SOX2) and 5 miRNAs (hsa-mir-103a-3p, hsa-mir-107, hsa-mir-16-5p, hsa-mir-34a-5p, and hsa-mir-23b-3p) as the transcriptional and posttranscriptional regulators of hHubGs. Three hHubGs (JUN, TP53, and ESR1) also act as TFs. As for example, the JUN inhibits to both TFs KLF4 and SOX2, TP53 activates to the TF POU5F1 and inhibits the TFs MYC and SOX2, and ESR1 activates to the TFs KLF4. Finally, this study recommended 10 candidate drug molecules (SORAFENIB, AMG-900, CHEMBL1765740, ENTRECTINIB, MK-6592, masitinib, GSK2126458, TG-02, YM201636, and PAZOPANIB) for the treatment against BC, where 1 drug (ENTRECTINIB) already approved by the Japanese drug administration authority, 3 drug molecules SORAFENIB, GSK2126458, and PAZOPANIB are under the clinical trial for BC treatment and the rest 6 molecules (AMG-900, CHEMBL1765740, MK-6592, masitinib, TG-02, and YM201636) require further experimental validation in wet-laboratory although they received support by some independent computational studies for BC treatment. Therefore, the findings of this study might be played a vital role for taking a proper treatment plan against BC.

Footnotes

Acknowledgements

The authors thank Dr Li at Taicang Affiliated Hospital of Soochow University for her kind suggestions and comments on our article.

Author Contributions

Conceptualization: MSA, GW, and MNHM; methodology: MSA, AS, MKK, GW, and MNHM; software: MSA, AS, and MKK; validation: MSA, GW, and MNHM; formal analysis: MSA, AS, and MKK; investigation: MSA., GW, and MNHM; data curation: MSA; writing—original draft preparation: MSA; writing—review and editing: MSA, AS, MKK, GW, and MNHM; visualization: MSA; supervision: GW and MNHM; project administration: GW and MNHM; funding acquisition: GW and MNHM. All authors have read and agreed to the published version of the article.

Declaration of conflicting interests:

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding:

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Talent Program of Taicang Health Commission (2022), the National Natural Science Foundation of China (grant no. 32271039), the Interdisciplinary Basic Frontier Innovation Program of Suzhou Medical College of Soochow University (grant no. YXY2303023), and Rajshahi University Research Project (A-289/5/52/RU/Science-24/2022-2023).

Data Availability

We collected HubG sets and drug molecules as molecular metadata from different published articles up to December 31, 2022, through the online search strategy with the keywords “breast cancer [ti], gene expression, RNA-Seq, scRNA-Seq, differentially expressed genes (DEGs) and core/key/hub genes” from the public sources including PubMed in the NCBI (https://pubmed.ncbi.nlm.nih.gov/), Google Scholar (https://scholar.google.com/), and Google (https://www.google.com/). We found a total of 297 articles based on the inclusion and exclusion criteria as shown in Figure S1. Finally, we selected 74 articles to collect the necessary molecular metadata. We collected 74 HubG sets from the selected 74 articles. We found a total of 259 unique HubGs from those of 74 HubG sets and used them for further investigation in this study (see Supplemental Table S1). The necessary meta-drug agents were collected from the online databases GSCALite (http://bioinfo.life.hust.edu.cn/web/GSCALite/) and Drug Gene Interaction Database (DGIdb) and published articles (Supplemental Table S2).

ORCID iD

Md Shahin Alam

Supplemental Material

Supplemental material for this article is available online.

References

Hanahan

Weinberg

RA.

Hallmarks of cancer: the next generation. Cell. 2011;144:646-674.

Gray

Collins

Genome changes and gene expression in human solid tumors. Carcinogenesis. 2000;21:443-452.

Bray

Ferlay

Soerjomataram

Siegel

Torre

Jemal

Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68:394-424.

Sung

Ferlay

Siegel

, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71:209-249.

Modern Cancer Hospital Guangzhou. Treatment effect and survival rate of breast cancer. Published 2017. Accessed August 14, 2024. https://www.asiancancer.com/breast-cancer-center/5656.html

Tesfaw

Teshale

Muluneh

EK.

Assessing the incidence, epidemiological description and associated risk factors of breast cancer in western Amhara, Ethiopia. Breast Cancer Manag. 2020;9:BMT47.

Ahmed

Reza

Sarker

, et al. Identification of host transcriptome-guided repurposable drugs for SARS-CoV-1 infections and their validation with SARS-CoV-2 infections by using the integrated bioinformatics approaches. PLoS ONE. 2022;17:e0266124.

Reza

Harun-Or-Roshid

Islam

, et al. Bioinformatics screening of potential biomarkers from mRNA expression profiles to discover drug targets and agents for cervical cancer. Int J Mol Sci. 2022;23:3968.

Islam

Rahman

Gov

, et al. Drug targeting and biomarkers in head and neck cancers: insights from systems biology analyses. OMICS. 2018;22:422-436.

10.

Mosharaf

Reza

Kibria

, et al. Computational identification of host genomic biomarkers highlighting their functions, pathways and regulators that influence SARS-CoV-2 infections and drug repurposing. Sci Rep. 2022;12:4279.

11.

Alam

Sultana

Sun

, et al. Bioinformatics and network-based screening and discovery of potential molecular targets and small molecular drugs for breast cancer. Front Pharmacol. 2022;13:942126.

12.

Alam

Rahaman

Sultana

Wang

Mollah

MNH

. Statistics and network-based approaches to identify molecular mechanisms that drive the progression of breast cancer. Comput Biol Med. 2022;145:105508.

13.

Langedijk

Mantel-Teeuwisse

Slijkerman

Schutjens

MH.

Drug repositioning and repurposing: terminology and definitions in literature. Drug Discov Today. 2015;20:1027-1034.

14.

Hua

Dai

, et al. Drug repositioning: Progress and challenges in drug discovery for various diseases. Eur J Med Chem. 2022;234:114239.

15.

Turanli

Grøtli

Boren

, et al. Drug repositioning for effective prostate cancer treatment. Front Physiol. 2018;9:500.

16.

Hopkins

AL.

Network pharmacology. Nat Biotechnol. 2007;25:1110-1111.

17.

Song

Zhang

Dai

Wang

Mechanism of action of celastrol against rheumatoid arthritis: a network pharmacology analysis. Int Immunopharmacol. 2019;74:105725.

18.

Dong

Zhong

Peng

Parallelization of molecular docking: a review. Curr Top Med Chem. 2018;18:1015-1028.

19.

Alam

Sultana

Wang

Haque Mollah

MN.

Gene expression profile analysis to discover molecular signatures for early diagnosis and therapies of triple-negative breast cancer. Front Mol Biosci. 2022;9:1049741.

20.

Alam

Sultana

Reza

Amanullah

Kabir

Mollah

MNH

. Integrated bioinformatics and statistical approaches to explore molecular biomarkers for breast cancer diagnosis, prognosis and therapies. PLoS ONE. 2022;17:e0268967.

21.

Mosharaf

Kibria

Hossen

, et al. Meta-data analysis to explore the hub of the hub-genes that influence SARS-CoV-2 infections highlighting their pathogenetic processes and drugs repurposing. Vaccines (Basel). 2022;10:1248.

22.

Reza

Hossen

Harun-Or-Roshid

Siddika

Kabir

Mollah

MNH

. Metadata analysis to explore hub of the hub-genes highlighting their functions, pathways and regulators for cervical cancer diagnosis and therapies. Discov Oncol. 2022;13:79.

23.

Liu

Xia

Han

Zhang

Guo

AY.

GSCALite: a web server for gene set cancer analysis. Bioinformatics. 2018;34:3771-3772.

24.

Wagner

Coffman

Ainscough

, et al.DGIdb 2.0: mining clinically relevant drug-gene interactions. Nucleic Acids Res. 2016;44:D1036-D1044.

25.

Peng

Jin

Circular RNA hsa_circ_0000376 participates in tumorigenesis of breast cancer by targeting miR-1285-3p. Technol Cancer Res Treat. 2020;19:1533033820928471.

26.

Hao

Liu

Ding

, et al. Identification of hub genes and small molecule therapeutic drugs related to breast cancer with comprehensive bioinformatics analysis. PeerJ. 2020;8:e9946.

27.

Sultana

Alam

Liu

, et al. Single-cell RNA-seq analysis to identify potential biomarkers for diagnosis, and prognosis of non-small cell lung cancer by using comprehensive bioinformatics approaches. Transl Oncol. 2023;27:101571.

28.

Singla

Sultana

Alam

Shen

Regulation of pain genes-capsaicin vs resiniferatoxin: reassessment of transcriptomic data. Front Pharmacol. 2020;11:551786.

29.

Szklarczyk

Gable

Lyon

, et al. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019;47:D607-D613.

30.

Shannon

Markiel

Ozier

, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13:2498-2504.

31.

Chandrashekar

Karthikeyan

Korla

, et al. UALCAN: an update to the integrated cancer data analysis platform. Neoplasia. 2022;25:18-27.

32.

Maubant

Tesson

Maire

, et al. Transcriptome analysis of Wnt3a-treated triple-negative breast cancer cells. PLoS ONE. 2015;10:e0122333.

33.

Pedraza

Gomez-Capilla

Escaramis

, et al. Gene expression signatures in breast cancer distinguish phenotype characteristics, histologic subtypes, and tumor invasiveness. Cancer. 2010;116:486-496.

34.

Chang

Gao

Hou

, et al. High-throughput sequencing reveals the disruption of methylation of imprinted gene in induced pluripotent stem cells. Cell Res. 2014;24:293-306.

35.

Kuleshov

Jones

Rouillard

, et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 2016;44:W90-97.

36.

Dennis

Jr Sherman

Hosack

, et al. DAVID: database for annotation, visualization, and integrated discovery. Genome Biol. 2003;4:P3.

37.

Zhou

Soufan

Ewald

Hancock

REW

Basu

Xia

NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis. Nucleic Acids Res. 2019;47:W234-W241.

38.

Berman

Battistuz

Bhat

, et al. The protein data bank. Acta Crystallogr D Biol Crystallogr. 2002;58:899-907.

39.

Waterhouse

Bertoni

Bienert

, et al. SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Res. 2018;46:W296-W303.

40.

Kim

Chen

Cheng

, et al. PubChem 2019 update: improved access to chemical data. Nucleic Acids Res. 2019;47:D1102-D1109.

41.

DeLano

Bromberg

PyMOL user’s guide. Published 2004. Accessed August 14, 2024. http://pymol.sourceforge.net/newman/userman.pdf

42.

Trott

Olson

AJ.

AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem. 2010;31:455-461.

43.

Dallakyan

Olson

AJ.

Small-molecule library screening by docking with PyRx. Methods Mol Biol. 2015;1263:243-250.

44.

Pettersen

Goddard

Huang

, et al. UCSF Chimera—a visualization system for exploratory research and analysis. J Comput Chem. 2004;25:1605-1612.

45.

Land

Humble

MS.

YASARA: a tool to obtain structural guidance in biocatalytic investigations. Methods Mol Biol. 2018;1685:43-67.

46.

Dickson

Madej

Skjevik

, et al. Lipid14: the amber lipid force field. J Chem Theory Comput. 2014;10:865-879.

47.

Guvench

MacKerell

Jr.

Comparison of protein force fields for molecular dynamics simulations. Methods Mol Biol. 2008;443:63-88.

48.

Ramirez

Caballero

Is it reliable to take the molecular docking top scoring position as the best solution without considering available structural data?

Molecules. 2018;23:1038.

49.

Harrach

Drossel

Structure and dynamics of TIP3P, TIP4P, and TIP5P water near smooth and atomistic walls of different hydroaffinity. J Chem Phys. 2014;140:174501.

50.

Krieger

Vriend

New ways to boost molecular dynamics simulations. J Comput Chem. 2015;36:996-1007.

51.

Miyamoto

Kollman

PA.

Settle: an analytical version of the SHAKE and RATTLE algorithm for rigid water models. J Comput Chem. 1992;13:952-962.

52.

Hess

Bekker

Berendsen

HJC

Fraaije

JGEM

. LINCS: a linear constraint solver for molecular simulations. J Comput Chem. 1998;18:1463-1472.

53.

Krieger

Koraimann

Vriend

Increasing the precision of comparative models with YASARA NOVA—a self-parameterizing force field. Proteins. 2002;47:393-402.

54.

Mitra

Dash

Structural dynamics and quantum mechanical aspects of shikonin derivatives as CREBBP bromodomain inhibitors. J Mol Graph Model. 2018;83:42-52.

55.

Ding

Zou

Wang

CCNB1 is a prognostic biomarker for ER+ breast cancer. Med Hypotheses. 2014;83:359-364.

56.

Yuan

Zheng

Liao

Overexpression of CCNE1 confers a poorer prognosis in triple-negative breast cancer identified by bioinformatic analysis. World J Surg Oncol. 2021;19:86.

57.

Deng

Wang

Identification of potential crucial genes and key pathways in breast cancer using bioinformatic analysis. Front Genet. 2019;10:695.

58.

Wang

Zhang

Huang

Integrated bioinformatics analysis reveals key candidate genes and pathways in breast cancer. Mol Med Rep. 2018;17:8091-8100.

59.

Jin

Huang

Shao

, et al. Integrated bioinformatics analysis to identify 15 hub genes in breast cancer. Oncol Lett. 2019;18:1023-1034.

60.

Lin

, et al. Identification of potential key genes for HER-2 positive breast cancer based on bioinformatics analysis. Medicine (Baltimore). 2020;99:e18445.

61.

Zhou

Chen

, et al. Significant prognostic values of differentially expressed-aberrantly methylated hub genes in breast cancer. J Cancer. 2019;10:6618-6634.

62.

Wei

Wang

Zhou

, et al. Bioinformatics analysis on enrichment analysis of potential hub genes in breast cancer. Transl Cancer Res. 2021;10:2399-2408.

63.

Luo

, et al. The prognostic significance of topoisomerase II alpha protein in early stage luminal breast cancer. BMC Cancer. 2018;18:331.

64.

Wang

Zhang

Jiang

Zhao

Teng

Identification of novel biomarkers in breast cancer via integrated bioinformatics analysis and experimental validation. Bioengineered. 2021;12:12431-12446.

65.

Shi

Shen

Liu

Yin

Identifying biomarkers to predict the progression and prognosis of breast cancer by weighted gene co-expression network analysis. Front Genet. 2020;11:597888.

66.

Ghafouri-Fard

Khoshbakht

Hussen

, et al. A review on the role of cyclin-dependent kinases in cancers. Cancer Cell Int. 2022;22:325.

67.

Wei

Wang

, et al. Identification of hub genes in triple-negative breast cancer by integrated bioinformatics analysis. Gland Surg. 2021;10:799-806.

68.

Liu

, et al. Identification of candidate biomarkers correlated with the pathogenesis and prognosis of breast cancer via integrated bioinformatics analysis. Medicine (Baltimore). 2020;99:e23153.

69.

Kim

In silico analysis of differentially expressed genesets in metastatic breast cancer identifies potential prognostic biomarkers. World J Surg Oncol. 2021;19:188.

70.

Zhang

Gao

, et al. Microarray based analysis of gene regulation by mesenchymal stem cells in breast cancer. Oncol Lett. 2017;13:2770-2776.

71.

Yuan

Zhang

Bie

, et al. Integrative analysis of lncRNAs and miRNAs with coding RNAs associated with ceRNA crosstalk network in triple negative breast cancer. Onco Targets Ther. 2017;10:5883-5897.

72.

Dashti

Taheri

Ghafouri-Fard

An in silico method leads to recognition of hub genes and crucial pathways in survival of patients with breast cancer. Sci Rep. 2020;10:18770.

73.

Gao

Han

CCNA2 is a prognostic biomarker for ER+ breast cancer and tamoxifen resistance. PLoS ONE. 2014;9:e91771.

74.

Dong

Zhang

Wei

, et al. Bioinformatic analysis of differential expression and core GENEs in breast cancer. Int J Clin Exp Pathol. 2018;11:1146-1156.

75.

Chen

Liu

Cen

, et al. KEGG-expressed genes and pathways in triple negative breast cancer: protocol for a systematic review and data mining. Medicine (Baltimore). 2020;99:e19986.

76.

Zhang

Qiu

Bai

Chen

HOTAIR facilitates endocrine resistance in breast cancer through ESR1/miR-130b-3p axis: comprehensive analysis of mRNA-miRNA-lncRNA network. Int J Gen Med. 2021;14:4653-4663.

77.

Zhao

Zhou

Qin

Estrogen receptor 1 and progesterone receptor are distinct biomarkers and prognostic factors in estrogen receptor-positive breast cancer: evidence from a bioinformatic analysis. Biomed Pharmacother. 2020;121:109647.

78.

Wang

Identification of key genes involved in tamoxifen-resistant breast cancer using bioinformatics analysis. Transl Cancer Res. 2021;10:5246-5257.

79.

Takeshita

Yan

Peng

, et al. Transcriptomic and functional pathway features were associated with survival after pathological complete response to neoadjuvant chemotherapy in breast cancer. Am J Cancer Res. 2020;10:2555-2569.

80.

Zheng

Wang

Molecular mechanisms of breast cancer metastasis by gene expression profile analysis. Mol Med Rep. 2017;16:4671-4677.

81.

Yuan

Jiang

, et al. Identification of differentially expressed lncRNAs and mRNAs in luminal-B breast cancer by RNA-sequencing. BMC Cancer. 2019;19:1171.

82.

de Araújo

da Luz

FAC

da Costa Marinho

, et al. Epidermal growth factor receptor (EGFR) expression in the serum of patients with triple-negative breast carcinoma: prognostic value of this biomarker. Ecancermedicalscience. 2022;16:1431.

83.

Wang

, et al. Integrated bioinformatics data analysis reveals prognostic significance of SIDT1 in triple-negative breast cancer. Onco Targets Ther. 2019;12:8401-8410.

84.

Amjad

Asnaashari

Sokouti

Dastmalchi

Systems biology comprehensive analysis on breast cancer for identification of key gene modules and genes associated with TNM-based clinical stages. Sci Rep. 2020;10:10816.

85.

Gao

Liu

, et al. Identification of the key pathways and genes involved in HER2-positive breast cancer with brain metastasis. Pathol Res Pract. 2019;215:152475.

86.

Bai

Luo

Zhang

Microarray data analysis reveals gene expression changes in response to ionizing radiation in MCF7 human breast cancer cells. Hereditas. 2020;157:37.

87.

Bhar

Haubrock

Mukhopadhyay

Maulik

Bandyopadhyay

Wingender

Coexpression and coregulation analysis of time-series gene expression data in estrogen-induced breast cancer cell. Algorithms Mol Biol. 2013;8:9.

88.

Wang

Zhu

Qiu

Lin

Systematic identification of the key candidate genes in breast cancer stroma. Cell Mol Biol Lett. 2018;23:44.

89.

Peng

Xia

Zheng

Integrated analysis of differentially expressed genes and pathways in triplenegative breast cancer. Mol Med Rep. 2017;15:1087-1094.

90.

Wang

Zhang

, et al. Aberrant Cyclin D1 splicing in cancer: from molecular mechanism to therapeutic modulation. Cell Death Dis. 2023;14:244.

91.

Krum

Miranda

Lin

Lane

TF.

BRCA1 associates with processive RNA polymerase II. J Biol Chem. 2003;278:52012-52020.

92.

Lashen

Toss

Katayama

Gogna

Mongan

Rakha

EA.

Assessment of proliferation in breast cancer: cell cycle or mitosis? An observational study. Histopathology. 2021;79:1087-1098.

93.

Mester

Redeuilh

Proliferation of breast cancer cells: regulation, mediators, targets for therapy. Anticancer Agents Med Chem. 2008;8:872-885.

94.

Ragoussis

Regulators of asymmetric cell division in breast cancer. Trends Cancer. 2018;4:798-801.

95.

Tulchin

Chambon

Juan

, et al. BRCA1 protein and nucleolin colocalize in breast carcinoma tissue and cancer cell lines. Am J Pathol. 2010;176:1203-1214.

96.

Adam

Boyd

Tyson

, et al. Comprehensive proteomic analysis of breast cancer cell membranes reveals unique proteins with potential roles in clinical cancer. J Biol Chem. 2003;278:6482-6489.

97.

Fang

Wang

Gao

Chen

Overexpression of CXXC5 is a strong poor prognostic factor in ER+ breast cancer. Oncol Lett. 2018;16:395-401.

98.

Song

Huang

, et al. Diagnostic and prognostic role of NR3C4 in breast cancer through a genomic network understanding. Pathol Res Pract. 2021;217:153310.

99.

Dolai

Liu

Molloy

MP.

Quantitative chemical proteomics in small-scale culture of phorbol ester stimulated basal breast cancer cells. Proteomics. 2011;11:2683-2692.

100.

Liu

Integrative bioinformatics analysis of KPNA2 in six major human cancers. Open Med (Wars). 2021;16:498-511.

101.

Tang

Zhao

Zhang

Wang

Identification of hub genes to regulate breast cancer metastasis to brain by bioinformatics analyses. J Cell Biochem. 2019;120:9522-9531.

102.

Tang

, et al. The role of upregulated miR-375 expression in breast cancer: an in vitro and in silico study. Pathol Res Pract. 2020;216:152754.

103.

Kane

Farrell

Saber

, et al. Sorafenib for the treatment of advanced renal cell carcinoma. Clin Cancer Res. 2006;12:7271-7278.

104.

Bianchi

Loibl

Zamagni

, et al. Phase II multicenter, uncontrolled trial of sorafenib in patients with metastatic breast cancer. Anticancer Drugs. 2009;20:616-624.

105.

Moreno-Aspitia

Morton

Hillman

Lingle

Rowland

KM.

Phase II trial of sorafenib in patients with metastatic breast cancer previously exposed to anthracyclines or taxanes: North Central Cancer Treatment Group and Mayo Clinic Trial N0336. J Clin Oncol. 2009;27:11-15.

106.

Wilhelm

Adnane

Newell

Villanueva

Llovet

Lynch

Preclinical overview of sorafenib, a multikinase inhibitor that targets both Raf and VEGF and PDGF receptor tyrosine kinase signaling. Mol Cancer Ther. 2008;7:3129-3140.

107.

Carducci

Shaheen

Markman

, et al. A phase 1, first-in-human study of AMG 900, an orally administered pan-Aurora kinase inhibitor, in adult patients with advanced solid tumors. Invest New Drugs. 2018;36:1060-1071.

108.

Juan

Bush

, et al. AMG 900, a potent inhibitor of aurora kinases causes pharmacodynamic changes in p-Histone H3 immunoreactivity in human tumor xenografts and proliferating mouse tissues. J Transl Med. 2014;12:307.

109.

Jiang

Chen

Entrectinib, a new multi-target inhibitor for cancer therapy. Biomed Pharmacother. 2022;150:112974.

110.

Liu

Offin

Harnicar

Drilon

Entrectinib: an orally available, selective tyrosine kinase inhibitor for the treatment of NTRK, ROS1, and ALK fusion-positive solid tumors. Ther Clin Risk Manag. 2018;14:1247-1252.

111.

European Medicines Agency. Refusal of the marketing authorisation for Masipro (masitinib). Published 2017. Accessed August 14, 2024. https://www.ema.europa.eu/en/documents/smop-initial/questions-answers-refusal-marketing-authorisation-masipro-masitinib_en.pdf

112.

European Commission. Union register of refused medicinal products for human use. Published 2018. Accessed August 14, 2024. https://ec.europa.eu/health/documents/community-register/html/ho26580.htm

113.

Munster

Aggarwal

Hong

, et al. First-in-human phase I study of GSK2126458, an oral pan-class I phosphatidylinositol-3-kinase inhibitor, in patients with advanced solid tumor malignancies. Clin Cancer Res. 2016;22:1932-1939.

114.

Feng

Jiang

Hao

GSK2126458 has the potential to inhibit the proliferation of pancreatic cancer uncovered by bioinformatics analysis and pharmacological experiments. J Transl Med. 2021;19:373.

115.

William

Lee

Goh

, et al. Discovery of kinase spectrum selective macrocycle (16E)-14-methyl-20-oxa-5,7,14,26-tetraazatetracyclo[19.3.1.1(2,6).1(8,12)]heptaco sa-1(25),2(26),3,5,8(27),9,11,16,21,23-decaene (SB1317/TG02), a potent inhibitor of cyclin-dependent kinases (CDKs), Janus kinase 2 (JAK2), and fms-like tyrosine kinase-3 (FLT3) for the treatment of cancer. J Med Chem. 2012;55:169-196.

116.

Administration USFaD. FDA grants orphan drug designation to zotiraciclib for the treatment of glioma. Published 2020. Accessed August 14, 2024. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0268967

117.

Zhang

Hei

, et al. CDK inhibitors in cancer therapy, an overview of recent development. Am J Cancer Res. 2021;11:1913-1935.

118.

Hou

Niu

, et al. Inhibition of PIKfyve using YM201636 suppresses the growth of liver cancer via the induction of autophagy. Oncol Rep. 2019;41:1971-1979.

119.

Taylor

Chia

Dent

, et al. A phase II study of pazopanib in patients with recurrent or metastatic invasive breast carcinoma: a trial of the Princess Margaret Hospital phase II consortium. Oncologist. 2010;15:810-818.

120.

Keisner

Shah

SR.

Pazopanib: the newest tyrosine kinase inhibitor for the treatment of advanced or metastatic renal cell carcinoma. Drugs. 2011;71:443-454.