Screening and identification of key biomarkers associated with endometriosis using bioinformatics and next-generation sequencing data analysis

Abstract

Background

Endometriosis is a common cause of endometrial-type mucosa outside the uterine cavity with symptoms such as painful periods, chronic pelvic pain, pain with intercourse and infertility. However, the early diagnosis of endometriosis is still restricted. The purpose of this investigation is to identify and validate the key biomarkers of endometriosis.

Methods

Next-generation sequencing dataset GSE243039 was obtained from the Gene Expression Omnibus database, and differentially expressed genes (DEGs) between endometriosis and normal control samples were identified. After screening of DEGs, gene ontology (GO) and REACTOME pathway enrichment analyses were performed. Furthermore, a protein–protein interaction (PPI) network was constructed and modules were analyzed using the Human Integrated Protein–Protein Interaction rEference database and Cytoscape software, and hub genes were identified. Subsequently, a network between miRNAs and hub genes, and network between TFs and hub genes were constructed using the miRNet and NetworkAnalyst tool, and possible key miRNAs and TFs were predicted. Finally, receiver operating characteristic curve analysis was used to validate the hub genes.

Results

A total of 958 DEGs, including 479 upregulated genes and 479 downregulated genes, were screened between endometriosis and normal control samples. GO and REACTOME pathway enrichment analyses of the 958 DEGs showed that they were mainly involved in multicellular organismal process, developmental process, signaling by GPCR and muscle contraction. Further analysis of the PPI network and modules identified 10 hub genes, including vcam1, snca, prkcb, adrb2, foxq1, mdfi, actbl2, prkd1, dapk1 and actc1. Possible target miRNAs, including hsa-mir-3143 and hsa-mir-2110, and target TFs, including tcf3 (transcription factor 3) and clock (clock circadian regulator), were predicted by constructing a miRNA-hub gene regulatory network and TF-hub gene regulatory network.

Conclusions

This investigation used bioinformatics techniques to explore the potential and novel biomarkers. These biomarkers might provide new ideas and methods for the early diagnosis, treatment and monitoring of endometriosis.

Background

Endometriosis is one of the most important chronic inflammatory diseases and has become the key cause of serious reproductive and general health condition [1]. Endometriosis is characterized by the presence of endometrial-type mucosa outside the uterine cavity [2]. The outbreak of their first period occurs through menopause, disregarding of ethnic origin or social status [2]. The clinical incidence of endometriosis is high, and its main features include dysmenorrhea, dyspareunia, chronic pelvic pain, irregular uterine bleeding and infertility [3], which places a great burden on the economy of health and reduces quality of life in worldwide; 10% of women of reproductive age are diagnosed with endometriosis each year [4]. These patients might have various complications including gynecological cancer (ovarian, endometrial and cervical cancers) [5], polycystic ovary syndrome [6], cardiovascular diseases [7], obesity [8], gestational diabetes mellitus [9], diabetes mellitus [10] and hypertension [11]. Endometriosis is recognized to be vulnerable on sex hormone estrogen, which rise the inflammation, growth and pain linked with the disease [6]. Studies have revealed that the progression of endometriosis is related to genetic risk factors [12] as well as environmental factors [13]. Because of this disorder complex pathogenesis, it is mainly treated by gynecological surgery [14], oral contraceptives [15], progestins [16], nonsteroidal anti-inflammatory drugs [17] and gonadotropin-releasing hormone agonists [18]. But these treatments have not been effective for longer period. Therefore, it is urgent to find specific molecular biomarkers for early assessment of the prognosis of patients with endometriosis, so as to further advancing the treatment and prognosis of patients. In current years, molecular biomarkers were demonstrated highly useful as clinical tools for endometriosis diagnosis and treatment [19].

The underlying complex molecular mechanisms in endometriosis pose a special challenge to daily clinical practice. Studies have found that genes including cyr61 [20], esr2 and cyp19a1 [21], hoxa10 [22], foxd3 [23], loxl1 and htra1 [24] can be used as an important marker for early diagnosis, prognosis and treatment of endometriosis. Studies have shown that signaling pathways including AKT and ERK signaling pathways [25], Wnt/β-catenin signaling pathway [26], PI3K-Akt-mTOR and MAPK signaling pathways [27], notch signaling pathway [28] and MAPK/ERK signal pathway [29] are involved in the progression of endometriosis. Taken together, current evidence suggests that the genes and signaling pathways are closely related to the progression of endometriosis.

The mechanisms of endometriosis at the molecular level are essentially for treating the disease. With the wide application of next-generation sequencing (NGS) technology, endometriosis-related genes have been widely identified, which is a key step in exploring the complex pathology of endometriosis and finding drugs that combat the illness. Numerous NGS data of gene expression have been published in public databases such as NCBI Gene Expression Omnibus (GEO) [https://www.ncbi.nlm.nih.gov/geo/] [30] during the past few years, and they are being increasingly used in bioinformatics and NGS data analysis to explore target genes or proteins associated in various diseases [31, 32].

Bioinformatics and network analysis of NGS data are an effective way to explore biomarkers in the pathogenesis of various diseases. Therefore, this investigation aimed to use bioinformatics analysis to identify hub genes and molecular pathways involved in endometriosis, to identify key diagnostic or therapeutic biomarkers. We obtained DEGs between endometriosis and normal control samples from GSE243039, a gene expression profile downloaded from the GEO database. Immediately after, we performed gene ontology (GO) and REACTOME pathway enrichment analysis on these DEGs. By constructing PPI networks, we screened for the significant modules and hub genes. We constructed miRNA-hub gene regulatory network and TF-hub gene regulatory network, and we screened for the miRNAs, TFs and hub genes. To validate that these hub genes can serve as molecular markers of endometriosis, we determined hub genes by using receiver operating characteristic curve (ROC) analysis. This investigation might offer better insight into potential molecular mechanisms to explore preventive and therapeutic strategies for endometriosis.

Methods

Next-generation sequencing (NGS) data source

The NGS dataset GSE243039 was obtained from the GEO database (Accession Date: 11/09/2023). The GSE243039 dataset included 20 endometriosis samples and 20 normal control samples. The platform used was the GPL24676 Illumina NovaSeq 6000 (Homo sapiens).

Identification of DEGs

The limma R/Bioconductor software package [33] was used to perform the identification of DEGs between endometriosis samples and normal control samples. We adjusted p-value to correct the false positive error caused by the multiple tests and determined it by the Benjamini and Hochberg method [34], which is the common tools to minimize the false discovery rate. The cutoff criteria were |logFC|> 1.304 (log2 fold change) for upregulated genes, |logFC|> 1.304 (log2 fold change) < − 1.2644 for downregulated genes and a adj.P.Val < 0.05. Thereafter, we used R packages “ggplot2” and “gplot” to show the DEGs with upregulated and downregulated expression in volcano plot and heatmap, respectively.

GO and pathway enrichment analyses of DEGs

GO enrichment analysis (http://www.geneontology.org) [35] (Accession Date: 26/02/2024) was frequently used to annotate the degree of gene function terms in DEGs, which included biological process (BP), cellular component (CC) and molecular function (MF). REACTOME (https://reactome.org/) (Accession Date: 26/02/2024) [36] pathway enrichment analysis was used to demonstrate enriched signaling pathways in DEGs. The g:Profiler (http://biit.cs.ut.ee/gprofiler/) (Accession Date: 26/02/2024) [37] was used to perform GO and REACTOME pathway enrichment analysis of DEGs. P < 0.05 was considered to represent statistical significance.

Construction of the PPI network and module analysis

To ensure the optimal graphical display of protein interactions of DEGs, Human Integrated Protein–Protein Interaction rEference (HIPIE) (https://cn.string-db.org/) [38] (Accession Date: 23/02/2024) was used to generate the PPI network. The software Cytoscape (version 3.10.1) (http://www.cytoscape.org/) [39] (Accession Date: 23/02/2024) was used to visualize the PPI network. The Network Analyzer in Cytoscape was utilized to calculate node degree [40], betweenness [41], stress [42] and closeness [43]. The PEWCC Cytoscape software plugin [44] was used to create modules in the PPI network of endometriosis.

Construction of the miRNA-hub gene regulatory network

The significant miRNAs were identified from miRNA-hub gene regulatory network analysis through the TarBase, miRTarBase, miRecords, miRanda (S mansoni only), miR2Disease, HMDD, PhenomiR, SM2miR, PharmacomiR, EpimiR, starBase, TransmiR, ADmiRE and TAM 2 via miRNet database (https://www.mirnet.ca/) [45] (Acession Date: 26/02/2024). This network was visualized with Cytoscape [39] (Accession Date: 26/02/2024), and the significant hub genes and miRNAs were selected via the Network Analyzer plugin in Cytoscape based on the degree connectivity.

Construction of the TF-hub gene regulatory network

The significant transcription factors (TFs) were identified from TF-hub gene regulatory network analysis through the CHEA via NetworkAnalyst database (https://www.networkanalyst.ca/) [46] (Acession Date: 01/05/2024). This network was visualized with Cytoscape [39] (Accession Date: 01/05/2024), and the significant hub genes and TFs were selected via the Network Analyzer plugin in Cytoscape based on the degree connectivity.

Receiver operating characteristic curve (ROC) analysis

ROC analysis was performed to predict the diagnostic effectiveness of biomarkers by pROC package of R software [47]. The area under the curve (AUC) value was utilized to determine the diagnostic effectiveness in discriminating endometriosis from normal control samples.

Results

Identification of DEGs

The DEGs were screened by “limma” package (|logFC|> 1.304 (log2 fold change) for upregulated genes and |logFC|> 1.304 (log2 fold change) < − 1.2644 for downregulated genes and adj.P.Val < 0.05). The GSE243039 dataset contained 958 DEGs, including 479 upregulated genes and 479 downregulated genes, and is listed in Table 1. The volcano plot (Fig. 1) was used to show the expression pattern of DEGs in endometriosis. The heatmap of the DEGs is shown in Fig. 2.

Table 1 The statistical metrics for key differentially expressed genes (DEGs)

Fig. 1

Volcano plot of differentially expressed genes. Genes with a significant change of more than twofold were selected. Green dot represented upregulated significant genes and red dot represented downregulated significant genes

Fig. 2

Heat map of differentially expressed genes. Legend on the top left indicates log fold change of genes. (A1 – A20 = Endometriosis samples; B1 – B20 = Normal control samples)

GO and pathway enrichment analyses of DEGs

GO enrichment and REACTOME pathway enrichment analyses were performed on the DEGs using the g:Profiler database. GO enrichment analysis covers three aspects: BP, CC and MF (Table 2). The upregulated genes were mainly related to multicellular organismal process, regulation of biological process, membrane, extracellular region, signaling receptor binding and molecular transducer activity, while the downregulated genes were mainly involved in developmental process, biological regulation, cell periphery, cytoplasm, molecular function regulator activity and calcium ion binding. The REACTOME pathway enrichment analysis showed that the genes upregulated genes in endometriosis were enriched in signaling by GPCR, extracellular matrix organization, muscle contraction and glycosaminoglycan metabolism (Table 3).

Table 2 The enriched GO terms of the up- and downregulated differentially expressed genes

Table 3 The enriched pathway terms of the up- and downregulated differentially expressed genes

Construction of the PPI network and module analysis

Considering the critical role of protein interactions in protein function, we used the HIPIE database and Cytoscape software to generate PPI network once we had identified the 958 DEGs. The results showed that there were dense regions in PPI, that is, genes closely related to endometriosis. A total of 4871 nodes and 8009 edges were selected to plot the PPI network (Fig. 3). The Network Analyzer plugin of Cytoscape was used to score each node gene by four selected algorithms, including node degree, betweenness, stress and closeness. Finally, we identified ten hub genes (vcam1, snca, prkcb, adrb2, foxq1, mdfi, actbl2, prkd1, dapk1 and actc1) and are listed in Table 4. The top two significant modules from PEWCC were selected for future analysis. Module 1 included 22 nodes and 41 edges (Fig. 4A). The functional enrichment analysis of genes in module 1 was conducted by g:Profiler. These genes were significantly enriched in multicellular organismal process and regulation of biological process. Module 2 included 8 nodes and 14 edges (Fig. 4B). The functional enrichment analysis of genes in module 2 was conducted by g:Profiler. These genes were significantly enriched in developmental process and biological regulation.

Fig. 3

PPI network of DEGs. Upregulated genes are marked in parrot green; downregulated genes are marked in red

Table 4 Topology table for up- and downregulated genes

Fig. 4

Modules selected from the PPI network. A The most significant module was obtained from PPI network with 22 nodes and 41 edges for upregulated genes B The most significant module was obtained from PPI network with 8 nodes and 14 edges for downregulated genes. Upregulated genes are marked in parrot green; downregulated genes are marked in red

Construction of the miRNA-hub gene regulatory network

We searched for target-regulated hub gene miRNAs using miRNet database and then used the results of this database. By constructing miRNA-hub gene regulatory network networks, we found 2495 nodes (miRNA: 2168; Hub Gene: 327) and 14,692 edges (Fig. 5). We identified 365 miRNAs (ex; hsa-mir-3143) targeting regulation of ccnd2, 102 mirnas (ex; hsa-mir-6888-5p) targeting regulation of vcam1, 89 mirnas (ex; hsa-mir-200a-3p) targeting regulation of ptprd, 88 mirnas (ex; hsa-mir-3122) targeting regulation of pdgfb, 81 mirnas (ex; hsa-mir-17-5p) targeting regulation of prkcb, 241 mirnas (ex; hsa-mir-2110) targeting regulation of igf2bp1, 77 mirnas (ex; hsa-mir-4432) targeting regulation of actc1, 53 mirnas (ex; hsa-mir-556-3p) targeting regulation of epb41l3, 48 mirnas (ex; hsa-mir-10b-5p) targeting regulation of dapk1 and 41 mirnas (ex; hsa-mir-1229-5p) targeting regulation of mdfi, and are listed in Table 5.

Fig. 5

Hub gene—miRNA regulatory network. The olive green color diamond nodes represent the key miRNAs; upregulated genes are marked in green; downregulated genes are marked in red

Table 5 MiRNA–hub gene and TF–hub gene topology table

Construction of the TF-hub gene regulatory network

We searched for target-regulated hub gene TFs using NetworkAnalyst database and then used the results of this database. By constructing TF-hub gene regulatory network networks, we found 520 nodes (TF: 198; Hub Gene: 322) and 8331 edges (Fig. 6). We identified 59 TFs (ex; tcf3) targeting regulation of ptch1, 42 tfs (ex; phc1) targeting regulation of ccnd2, 37 tfs (ex; nr1i2) targeting regulation of prkcb, 36 tfs (ex; hoxc9) targeting regulation of st8sia4, 34 tfs (ex; rnf2) targeting regulation of foxq1, 44 tfs (ex; clock) targeting regulation of dapk1, 38 tfs (ex; prdm14) targeting regulation of igf2bp1, 36 tfs (ex; smarca4) targeting regulation of krt18, 34 tfs (ex; trim28) targeting regulation of stx11 and 31 tfs (ex; htt) targeting regulation of epb41l3, and are listed in Table 5.

Fig. 6

Hub gene—TF regulatory network. The blue color triangle nodes represent the key TFs; upregulated genes are marked in dark green; downregulated genes are marked in dark red

Receiver operating characteristic curve (ROC) analysis

The ROC curve was used to evaluate the diagnostic value of hub genes. As shown in Fig. 7, the AUC values of vcam1, snca, prkcb, adrb2, foxq1, mdfi, actbl2, prkd1, dapk1 and actc1 in endometriosis were 0.904, 0.907, 0.903, 0.926, 0.901, 0.910, 0.923, 0.892, 0.895 and 0.898, respectively. Thus, the hub genes have good diagnostic efficiency in endometriosis and normal control samples.

Fig. 7

ROC curve analyses of hub genes. A VCAM1 B SNCA C PRKCB D ADRB2 E FOXQ1 F MDFI G ACTBL2 H PRKD1 I DAPK1 J ACTC1

Discussion

Endometriosis is a key cause of serious reproductive disorder in the female population and leads to a public health burden. Lack of early screening and diagnosis of endometriosis result in progressive worsening including dysmenorrhea, dyspareunia, chronic pelvic pain, irregular uterine bleeding and infertility. The advance stage of endometriosis seriously affects the recovery from female reproductive diseases. Therefore, it is necessary to identify potential novel biomarkers for early diagnosis and targeted therapy of endometriosis. With the advancement of bioinformatics methods and NGS technology, it has started to be widely applied to identify potential novel biomarkers. This investigation used NGS data to conduct bioinformatics analysis for identifying novel target genes and pathways involved in the occurrence and development of endometriosis. Bioinformatics and NGS data analysis might become to identify effective drugs for treating endometriosis in future.

In this investigation, we analyzed the endometriosis GSE243039 screened from the GEO database. It includes 20 normal control samples and 20 endometriosis samples. Compared to normal controls, we found 958 DEGs (including 479 upregulated genes and 479 downregulated genes). Research has shown that pcsk9 [48], cntn4 [49], sema3a [50], sfrp4 [51], mfap5 [52], bmp6 [53], cdh6 [54], piezo2 [55] and pkp2 [56] play an important role in the pathogenesis of inflammation. Some studies have shown that altered expression of genes includes pcsk9 [57], sema3a [58] and sfrp4 [59] promotes the pain. Studies have revealed that genes including pcsk9 [60], cntn4 [61], sema3a [62], ptgis [63], sfrp4 [64], mfap5 [65], cdh6 [66], gpc6 [67] and pkp2 [68] play a key role in ovarian cancer. A study indicates that genes including pcsk9 [69], sfrp4 [70] and bmp6 [71] have been identified in polycystic ovarian syndrome. Genes including pcsk9 [72], cntn4 [49], sema3a [73], sfrp4 [74], mfap5 [75], bmp6 [76], pde1c [77] and pkp2 [78] are altered expressed in cardiovascular diseases. The genes including pcsk9 [79], apcdd1 [80], sfrp4 [81], mfap5 [82] and pkp2 [83] have been identified to be involved in the development of obesity. Studies show that genes including pcsk9 [84] and sfrp4 [85] have been known to be involved in gestational diabetes mellitus. Recent reports have revealed that genes including pcsk9 [86], sema3a [87], sfrp4 [81] and bmp6 [88] play an important role in the pathogenesis of diabetes mellitus. Recent reports have revealed that genes including pcsk9 [89], sema3a [90], ptgis [91] and piezo2 [92] have a significant prognostic potential in hypertension. Studies have found that genes including adamts19 [93] and bmp6 [94] play an indispensable role in infertility. Previous studies have reported that the genes including sema3a [95], sfrp4 [96] and bmp6 [94] are a key regulator of endometriosis. Recently, increasing evidence demonstrated that genes including ptgis [97] and sfrp4 [98] might be potential therapeutic targets for endometrial cancer treatment. Studies have found that genes including sfrp4 [99] and mfap5 [100] are used as prognostic markers for cervical cancer. This result suggests that these genes might play a key role in the progression of endometriosis.

In this investigation, we identified enriched genes in GO terms and signaling pathways that might be utilized as diagnostic, prognostic and therapeutic targets in endometriosis. Signaling pathways including extracellular matrix organization [101], nervous system development [102], signal transduction [103], hemostasis [104], muscle contraction [105], signaling by retinoic acid [106] and diseases of glycosylation [107] were responsible for advancement of endometriosis. Recently, mounting researches have revealed that genes including l1cam [108], hsd17b2 [109], vcam1 [110], sox6 [111], fgf10 [112], mmp12 [113], ccr1 [114], prok1 [115], prl [116], timp3 [117], adamts9 [118], ndnf [119], lhcgr [120], pdgfb [121], ldlr [122], cd4 [123], foxl2 [124], trpa1 [125], adrb2 [126], plau 127], epcam [128], ucn2 [129], cyp1a1 [130], ntn1 [131], il15 [132], bmp2 [133], apoe [134], casp1 [135], abcg2 [136], ace [137], pgr [138], alpp [139], lpar4 [140], atrnl1 [141], hla-c [142], mmp3 [143], pdlim3 [144], nfasc [145], il33 [146], ngf [147], comp [148], fst [149], efemp1 [150], gata6 [151], tcf21 [152], ptgs2 [153], hoxc8 [154], akr1c3 [155], bdnf [119], epha3 [156], inhba [157], rap1gap [158], tlr3 [159], nox4 [160], tgfbi [161], igf2bp1 [162], dlx5 [163], vdr [164], fzd7 [165], id2 [166], tlr2 [167], il6 [168], gas6 [169], dusp2 [170], fgf7 [171], ccn2 [172], igfbp3 [173], chl1 [174], bgn [175], ntrk2 [176], slit2 [177], notch2 [178], lif [179], cd200 [180], bst2 [181], dysf [182], dapk1 [183], kiss1 [184], fpr1 [185] and trh [186] were vital for the onset and developmental process of endometriosis. A great number of studies have indicated that genes including l1cam [187], ajap1 [188], hsd17b2 [189], vcam1 [190], grp [191], aqp8 [192], wnt6 [193], fabp4 [194], sox6 [195], ntrk1 [196], cntn1 [197], mmp12 [198], lag3 [199], sox18 [200], ccr1 [201], flt1 [202], prdm1 [203], trpc3 [204], dkk2 [205], rnf157 [206], dhcr24 [207], bmp4 [208], prl [209], foxq1 [210], wnt5a [211], meox1 [212], dock4 [213], timp3 [214], adamts9 [215], ndnf [216], neto1 [217], cd24 [218], lhcgr [219], scd [220], pdgfb [221], mmrn1 [222], ldlr [223], cd4 [224], foxl2 [225], trpa1 [226], epha5 [227], tox [228], cst4 [229], rspo3 [230], map2k6 [231], nes [232], tmem119 [233], padi2 [234], mmp8 [235], kdr [236], adrb2 [237], mgat3 [238], ptprc [239], pitx1 [240], kl [241], plau [242], znf365 [243], pik3r3 [244], sox8 [245], ccnd2 [246], crabp2 [247], pcdh9 [248], epcam [249], clec14a [250], cyp1a1 [251], ntn1 [252], pdgfd [253], cldn3 [254], lepr [255], il15 [256], bmp2 [257], lama5 [258], ntng1 [259], krt19 [260], ros1 [261], apoe [262], ptch1 [263], itpka [264], casp1 [265], nid1 [266], abcg2 [267], ace [268], pgr [269], wls [270], klk3 [271], lrp1b [272], ly6k [273], alpp [274], prame [275], slco4a1 [276], egfl6 [277], gpbar1 [278], elmo1 [279], wnk2 [280], il2rb [281], diras2 [282], galnt14 [283], rtkn2 [284], atrnl1 [285], s100a4 [286], macc1 [287], mmp3 [288], col11a1 [289], cables1 [290], fgfr2 [291], il33 [292], ngf [293], foxc2 [294], comp [295], fst [296], sorbs2 [297], efemp1 [298], gata6 [299], tcf21 [300], ptgs2 [301], mtss1 [302], dact1 [303], hoxc8 [304], pitx2 [305], tnfsf10 [306], bdnf [307] krt7 [308], ndrg2 [309], eya2 [310], inhba [311], sgk1 [312], slc2a12 [313], dio3 [314], epb41l3 [315], tlr3 [316], angptl4 [317], ephb2 [318], fli1 [319], thbs1 [320], id3 [321], nox4 [322], tgfbi [323], igf2bp1 [324], sall4 [325], dlx5 [326], vdr [327], lzts1 [328], fzd7 [329], en2 [330], enc1 [331], ifne [332], tnnt1 [333], ankrd1 [334], sox9 [335], mgp [336], sulf1 [337], cyp24a1 [338], dnah11 [339], tlr2 [340], il6 [341], nppb [342], spink1 [343], gpc3 [344], ntrk3 [345], amigo2 [346], foxd1 [347], adam12 [348], dusp2 [349], usp2 [350], klf2 [351], sik1 [352], six1 [310], fgf7 [353], myh10 [354], igfbp3 [355], lyve1 [356], actbl2 [357], slit2 [358], actc1 [359], nnmt [360], chi3l1 [361], runx1 [362], nfib [363], notch2 [364], pgf [365], thbs2 [366], nav1 [367], nrg1 [368], plk2 [369], itgbl1 [370], cd200 [371], bst2 [372], kcnn3 [373], hmcn1 [374], veph1 [375], tfpi2 [376], sytl2 [377], ccdc80 [378], dapk1 [379], kiss1 [380], il20ra [381], has3 [382], has1 [382], mgst1 [383], fpr1 [384] and sh3rf2 [385] are closely associated with the onset and progression of ovarian cancer. A previous study reported that the genes including l1cam [386], hsd17b2 [387], grp [388], fabp4 [389], sox6 [390]. mmp12 [391], apod [392], lag3 [393], cst1 [394], flt1 [395], dhcr24 [396], prl [397], wnt5a [398], timp3 [399], cd24 [400], lhcgr [401], mmrn1 [402], cd4 [403], adamts5 [404], adamts1 [405], padi2 [406], mark1 [407], kl [408], plau [409], sox8 [410], crabp2 [411], ptprd [412], epcam [413], irx2 [414], sema3b [415], cyp1a1 [416], pdgfd [417], lepr [418], apoe [419], casp1 [420], mgll [421], nid1 [422], abcg2 [423], ace [424], pgr [425], hpse2 [426], lmtk3 [427], alpp [428], egfl6 [429], cacna2d3 [430], mctp1 [431], hkdc1 [432], s100a4 [433], macc1 [434], mmp3 [435], fgfr2 [436], il33 [437], foxc2 [438], itga7 [439], efemp1 [440], gata6 [441], bhlhe41 [442], tcf21 [443], gdf10 [444], nkx3-1 [445], akr1c3 [446], sgk1 [447], rap1gap [448], fli1 [449], nox4 [450], serpine2 [451], igsf9 [452], igf2bp1 [453], sall4 [454], vdr [455], celsr2 [456], enc1 [457], sox9 [458], cyp24a1 [338], il6 [459], gas6 [460], klf2 [461], six1 [462], igfbp3 [463], lyve1 [464], chl1 [465], bgn [466], slit2 [467], nrp2 [468], nnmt [469], runx1 [470], thbs2 [471], hspb7 [472], nrg1 [473], tfpi2 [474], dapk1 [183], has3 [475], has1 [475], steap1 [476] and mgst1 [477] play an important role in the pathophysiology of endometrial cancer. Study demonstrated that genes including vcam1 [478], aqp8 [479], l1cam [480], fabp4 [481], psg1 [482], sox6 [483], mmp12 [484], apod [485], lag3 [486], sox18 [487], flt1 [488], fabp5 [489], bmp4 [490], prl [491], foxq1 [492], wnt5a [493], frzb [494], cpe [495], ereg [496], ndnf [497], cd24 [498], scd [499], ldlr [500], cd4 [501], foxl2 [502], krt17 [503], nes [504], mgat3 [505], mark1 [506], kl [507], plau [508], epha7 [509], pik3r3 [510], ccnd2 [511], hecw1 [512], epcam [513], batf2 [514], cyp1a1 [515], mstn [516], il15 [517], syt7 [518], pak3 [519], krt19 [520], ros1 [521], cubn [522], ptch1 [523], casp1 [524], abcg2 [525], pgr [526], hpse2 [527], lrp1b [528], alpp [529], cyp2s1 [530], doc2b [531], msmo1 [532], sorcs1 [533], hla-c [534], s100a4 [535], macc1 [536], mmp3 [537], fgfr2 [538], il33 [539], ngf [540], foxc2 [541], sorbs2 [542], itga7 [543], efemp1 [544], gata6 [545], tcf21 [546], ptgs2 [547], mtss1 [548], dact1 [549], spint2 [550], nkx3-1 [551], hoxc8 [552], bdnf [497], ndrg2 [553], epha3 [554], eya2 [555], inhba [556], alpl [557], sgk1 [558], rap1gap [559], epb41l3 [560], tlr3 [561], angptl4 [562], ephb2 [563], fli1 [564], thbs1 [565], nox4 [566], tgfbi [567], igf2bp1 [568], sall4 [569], vdr [570], rarb [571], epha4 [572], enc1 [573], sox9 [574], sulf1 [575], tlr2 [576], il6 [577], gpc3 [578], ntrk3 [579], ccna1 [580], amigo2 [581], foxd1 [582], ccno [583], adam12 [584], rassf2 [585], hoxb7 [586], klf2 [587], sik1 [588], six1 [589], fgf7 [590], igfbp3 [591], chl1 [592], eppk1 [593], slit2 [594], flg [595], nrp2 [596], nnmt [597], chi3l1 [598], runx1 [599], apln [600], sema3c [601], notch2 [602], thbs2 [603], pnpla1 [604], bst2 [605], hmcn1 [606], ulbp1 [607], tfpi2 [608], dapk1 [609], kiss1 [610], fpr1 [611] and pik3ap1 [612] can participate in the occurrence and development of cervical cancer. The abnormal expression of genes including cbln2 [613], sdk1 [614], vcam1 [615], six2 [616], avpr1a [617], epha6 [618], fabp4 [619], psg1 [620], ano1 [621], sox6 [622], fgf10 [623], pla2g7 [624], mmp12 [625], adra1d [626], lag3 [627], flt1 [628], fabp5 [629], prdm1 [630], trpc3 [631], igsf3 [632], bmp4 [633], il1rl1 [634], prl [635], nefl [636], wnt5a [637], timp3 [638], ndnf [639], snap25 [640], cd24 [641], pdgfb [642], ldlr [643], cd4 [644], trpa1 [645], adamts1 [646], pde4b [647], nes [648], th [649], psg9 [650], cacna1d [651], mmp8 [652], adrb2 [653], kl [654], plau [655], ptprd [656], sema3b [657], ucn2 [658], cyp2j2 [659], cyp1a1 [660], atp1a2 [661], cldn3 [662], mstn [663], lepr [664], il15 [665], cacna1h [666], bmp2 [667], lama5 [668], ros1 [669], apoe [670], casp1 [671], pde9a [672], efnb2 [673], abcg2 [674], ace [675], pgr [676], slc35f3 [677], ica1 [678], alpp [679], trpc6 [680], gpbar1 [681], pnpla3 [682], hla-c [683], s100a4 [684], macc1 [685], mmp3 [686], gdnf [687], fgfr2 [688], il33 [689], ngf [690], pappa2 [691], comp [692], gata6 [693], acan [694], tcf21 [695], ptgs2 [696], pitx2 [697], akr1c3 [698], bdnf [699], sgk1 [700], tlr3 [701], angptl4 [702], fli1 [703], thbs1 [704], id3 [705], nox4 [706], pcsk1 [707], itgb1bp2 [708], wnk4 [709], dlx5 [710], vdr [711], epha4 [712], mgp [713], cyp24a1 [714], id2 [715], tlr2 [716], il6 [717], nppb [718], gas6 [719], f11r [720], foxd1 [721], adam12 [722], ncam1 [723], usp2 [724], klf2 [725], sik1 [726], fgf7 [727], igfbp3 [728], bgn [729], ntrk2 [730], nnmt [731], chi3l1 [732], runx1 [733], apln [734], stox2 [735], kcnq4 [736], notch2 [737], pgf [738], thbs2 [739], pdlim5 [740], prdm6 [741], htr6 [742], nrg1 [7 43], cd200 [744], bst2 [745], kcnn3 [746], slc2a5 [747], tfpi2 [748], dysf [749], ccdc80 [750], dapk1 [751], kiss1 [752], slc4a4 [753], steap2 [754], sorbs1 [755], ackr2 [756], fpr1 [757], gpr143 [758] and trh [759] contributes to the progression of hypertension. Study showed that the genes including robo2 [760], vcam1 [761], grp [762], fabp4 [763], ano1 [764], sox6 [765], tfap2c [766], ramp3 [767], pla2g7 [768], mmp12 [769], faim2 [770], apod [771], lag3 [772], sox18 [773], f2rl2 [774], ccr1 [775], flt1 [776], fabp5 [629], trpc3 [777]. thsd7a [778], dkk2 [779], prkcb [780], dhcr24 [781], pde3b [782], bmp4 [783], il1rl1 [784], mypn [785], plcg2 [786], prl [787], wnt5a [788], meox1 [789], timp3 [790], frzb [791], cpe [792], adamts9 [793], ndnf [794], pdgfb [795], pik3cg [796], ldlr [797], cd4 [798], trpa1 [799], f2rl3 [800], c1ql1 [801], adamts5 [802], pde4b [803], nes [804], th [805], mmp8 [806], kdr [807], adrb2 [808], ackr3 [809], ptprc [810], kl [811, 812], plau [813], ccnd2 [814], ptgs1 [815], insig1 [816], irx2 [817], siglec1 [818], ucn2 [819], cyp2j2 [820], cyp1a1 [821], astn2 [822], ntn1 [823], pdgfd [824], mstn [663], lepr [664], il15 [825], cacna1h [826], bmp2 [827], syt7 [828], zbtb46 [829], ros1 [830], apoe [831], cubn [832], rbm20 [833], casp1 [834], pde9a [835], abcg2 [836], hmgcr [837], ace [838], grem2 [839], palmd [840], lrp1b [841], alpp [842], trpc6 [843], gpbar1 [844], myzap [845], prodh [846], il2rb [847], cdhr3 [848], pnpla3 [849], fads1 [850], hla-c [851], s100a4 [852], mmp3 [853], pdlim3 [854], gdnf [855], fgfr2 [856], il33 [857], ngf [858], hapln1 [859], foxc2 [860], comp [861], fst [862], sorbs2 [863], itga7 [864], pln [865], gata6 [866], bhlhe41 [867], acan [868], tcf21 [869], ptgs2 [870], dact1 [871], pitx2 [872], akr1c3 [873], bdnf [874], ndrg2 [875], eya2 [876], sgk1 [877], rap1gap [878], dio3 [879], tlr3 [880], angptl4 [881], ephb2 [882], thbs1 [883], tnnt2 [884], nox4 [885], s1pr5 [886], serpine2 [887], pcsk1 [888], tgfbi [889], sall4 [890], eya4 [891], itgb1bp2 [708], vdr [892], gpc4 [893], celsr2 [894], epha4 [895], tnnt1 [896], ankrd1 [897], zfpm2 [898], sox9 [899], mgp [900], cyp24a1 [901], dnah11 [902], tlr2 [903], il6 [904], gas6 [905], gpc3 [906], ntrk3 [907], amigo2 [908], f11r [909], adam12 [722], ncam1 [910], usp2 [911], klf2 [912], sik1 [913], six1 [914], fgf7 [915], ccn2 [916], jcad [917], igfbp3 [918], lyve1 [919], prkd1 [920], bgn [921], eda 922], slit2 [923], actc1 [924], nrp2 [925], chi3l1 [926], runx1 [927], apln [928], myom2 [929], myoz1 [930], ppp1r13l [931], thbs2 [932], des [933], pdlim5 [934], hspb7 [935], nrg1 [936], plk2 [937], itgbl1 [938], cd200 [939], kcnn3 [940], kcnj2 [941], eva1a [942], tfpi2 [943], dysf [944], adap1 [945], ccdc80 [946], dapk1 [947], scn4b [948], esyt3 [949], abca8 [950], heg1 [951], fpr1 [952], sspn [953], adh1c [954], sirpa [955] and trh [956] might be related to the pathophysiology of cardiovascular diseases. A study showed genes including hsd17b2 [109], efna5 [957], mmp12 [958], prok1 [959], prl [960], nlrp2 [961], ndnf [962], mei4 [963], cd24 [964], lhcgr [965], cd4 [966], foxl2 [967], kdr [968], adrb2 [969], cyp1a1 [970], ntn1 [971], mstn [972], bmp2 [973], apoe [974], ace [975], pgr [976], grem2 [977], alpp [978], mmp3 [979], gdnf [980], fgfr2 [981], il33 [982], ngf [983], comp [984], cecr2 [985], fst [986], gata6 [987], ptgs2 [153], bdnf [988], sgk1 [989], angptl4 [990], thbs1 [991], id3 [992], nox4 [993], igf2bp1 [994], sall4 [995], vdr [996], sulf1 [997], tlr2 [998], il6 [999], gpc3 [1000], ccno [1001], igfbp3 [1002], chl1 [1003], ntrk2 [1004], slit2 [1005], apln [1006], notch2 [1007], pgf [1008], lif [1009], cd200 [1010], tfpi2 [1011], kiss1 [752] and trh [1012] are highly prone to infertility. A study indicated that activation of genes including vcam1 [761], stra6 [1013], coch [1014], grp [1015], aqp8 [1016], fabp4 [1017], ano1 [1018], sox6 [1019], tfap2c [1020], ntrk1 [1021], cntn1 [1022], fgf10 [1023], pla2g7 [768], mmp12 [1024], lcp1 [1025], snca [1026], apod [1027], lag3 [1028], ccr1 [1029], cst1 [1030], retreg1 [1031], flt1 [1032], fabp5 [1033], prdm1 [1034], trpc3 [1035], prok1 [1036], wnt16 [1037], f13a1 [1038], dhcr24 [1039], pde3b [1040], bmp4 [783], il1rl1 [1041], plcg2 [1042], prl [1043], foxq1 [1044], nefl [1045], wnt5a [1046], timp3 [1047], serpinb2 [1048], frzb [1049], nlrp2 [1050], cpe [1051], adamts9 [1052], npw [1053], ereg [1054], ndnf [1055], snap25 [1056], syt1 [1057], scd [1058], pdgfb [1059], ldlr [1060], cd4 [1061], gpr183 [1062], trpa1 [125], ptger4 [1063], adamts5 [1064], rspo3 [1065], krt17 [1066], adamts1 [1067], pde4b [1068], nes [804], sh2d2a [1069], th [1070], tmem119 [1071], mmp8 [1072], adrb2 [1073], ackr3 [1074], mgat3 [1075], tnfrsf9 [1076], txk [1077], kl [811], plau [1078], epha7 [1079], znf365 [1080], pik3r3 [1081], sox8 [1082], ccnd2 [1083], ptgs1 [1084], batf2 [1085], ucn2 [1086], cyp2j2 [1087], clec14a [1088], cyp1a1 [1089], ntn1 [1090], mstn [1091], lepr [1092], cd248 [1093], il15 [825], bmp2 [1094], ros1 [1095], mme [1096], apoe [1097], ptch1 [1098], casp1 [1099], mgll [1100], efnb2 [1101], abcg2 [1102], hmgcr [1103], ace [1104], pgr [1105], grem2 [839], rgs7 [1106], chst1 [1107], alpp [1108], trpc6 [1109], slco4a1 [276], cyp4b1 [1110], gpbar1 [844], elmo1 [1111], doc2b [1112], cd163l1 [1113], slco2a1 [1114], il2rb [1115], b4galnt2 [1116], slc37a2 [1117], pnpla3 [1118], fads1 [1119], hla-c [1120], st3gal5 [1121], s100a4 [1122], macc1 [1123], cort [1124], mmp3 [1125], gdnf [1126], lmo3 [1127], nfasc [1128], fgfr2 [1129], il33 [1130], ngf [1131], hapln1 [1132], gdf6 [1133], foxc2 [1134], comp [1135], fst [1136], itga7 [1137], gata6 [1138], acan [1139], tcf21 [1140], ptgs2 [1141], mtss1 [1142], dhrs3 [1143], nkx3-1 [1144], tnfsf10 [1145], bdnf [1146], ndrg2 [1147], epha3 [1148], pla2g5 [1149], mecom [1150], sgk1 [1151], tlr3 [1152], angptl4 [1153], ephb2 [1154], fli1 [1155], thbs1 [1156], mbp [1157], id3 [1158], nox4 [1159], s1pr5 [1160], pi16 [1161], igf2bp1 [1162], sall4 [1163], vldlr [1164], vdr [1165], fam20a [1166], epha4 [1167], ankrd1 [1168], sgca [1169], sox9 [1170], mgp [1171], cyp24a1 [1172], tlr2 [1173], il6 [1174], gas6 [1175], ntrk3 [1176], adam12 [1177], ncam1 [1178], myoc [1179], usp2 [1180], klf2 [1181], sik1 [1182], six1 [1183], fgf7 [1184], ccn2 [1185], igfbp3 [1186], lyve1 [1187], prkd1 [1188], bgn [1189], slit2 [1190], irx3 [1191], actc1 [1192], flg [1193], nrp2 [1194], nnmt [1195], chi3l1 [1196], runx1 [1197], nfib [1198], apln [1199], plp1 [1200], nav2 [1201], notch2 [1202], pgf [1203], thbs2 [1204], nrg1 [1205], lif [1206], plk2 [1207], nalcn [1208], cd200 [1209], kcnn3 [1210], eva1a [1211], tfpi2 [1212], dysf [1213], sytl2 [1214], tlr1 [1215], ccdc80 [946], dapk1 [947], kiss1 [1216], gem [1217], il20ra [1218], has3 [1219], has1 [1220], slc4a4 [1221], sirpb1 [1222], steap1 [1223], ackr2 [1224], fpr1 [1225], gng7 [1226], igfbpl1 [1227], pik3ap1 [1228], adh1c [1229], lxn [1230] and trh [1231] has been observed in inflammation. Studies have suggested that genes including vcam1 [1232], aqp8 [1233], fabp4 [1234], fabp5 [1235], bmp4 [1236], prl [960], wnt5a [1237], adamts9 [1238], ndnf [1239], lhcgr [1240], ldlr [1241], cd4 [1242], adamts5 [1243], map2k6 [1244], adamts1 [1245], pde4b [1246], th [1247], mmp8 [1248], adrb2 [969], kl [1249], epha7 [1250], ucn2 [1251], cyp1a1 [970], lepr [1252], il15 [1253], bmp2 [1254], apoe [1255], casp1 [1256], ace [1257], pgr [1258], grem2 [1259], sorcs1 [1260], hkdc1 [1261], fads1 [1262], s100a4 [1263], il33 [1243], ngf [1264], comp [1265], fst [1266], gata6 [1267], acan [1268], akr1c3 [1269], bdnf [1270], angptl4 [1271], nox4 [1272], vdr [1273], gpc4 [1274], tlr2 [1275], il6 [1276], igfbp3 [1277], tnik [1278], apln [1279], pgf [1280], nrg1 [1281], lif [1282], angptl1 [1283], kiss1 [1284], sorbs1 [1285] and trh [1286] can be used as important therapeutic targets for polycystic ovarian syndrome. A recent study found that genes including vcam1 [1287], stra6 [1288], aqp8 [1289], fabp4 [1290], sox6 [1291], ramp3 [1292], mmp12 [1293], faim2 [770], ccr1 [1294], ism1 [1295], flt1 [1296], fabp5 [1297], thsd7a [1298], sctr [1299], wnt16 [1300], prkcb [1301], pde3b [1302], il1rl1 [1041], prl [1303], wnt5a [1304], htr1b [1305], timp3 [1306], cpe [1051], ereg [1307], ndnf [1308], snap25 [1309], cd24 [1310], scd [1058], pdgfb [1311], ldlr [1312], cd4 [1313], trpa1 [1314], map2k6 [1315], pde4b [1316], th [1317], mmp8 [1318], adrb2 [1319], kl [1320], plau [1321], ptgs1 [1084], insig1 [1322], bmp8a [1323], ucn2 [1324], ntn1 [1090], pdgfd [1325], mstn [1326], lepr [1092], il15 [1327], bmp2 [1328], apoe [1329], casp1 [1330], mgll [1331], nid1 [1332], abcg2 [1333], ace [1334], pgr [1335], grem2 [1336], lrp1b [1337], alpp [1338], trpc6 [1339], egfl6 [1340], gpbar1 [1341], aif1l [1342], gpat3 [1343], sorcs1 [1260], slc37a2 [1344], fads1 [1345], acsl5 [1346], ptprn2 [1347], s100a4 [1348], macc1 [1349], cort [1350], mmp3 [1351], gdnf [1352], lmo3 [1353], cables1 [1354], il33 [1355], ngf [1356], foxc2 [1357], fst [1136], pln [1358], acan [1359], ptgs2 [1360], gdf10 [1361], cpne5 [1362], dgat2 [1363], bdnf [1364], rgs4 [1365], epha3 [1366], pla2g5 [1367], sgk1 [700], tlr3 [1368], angptl4 [1369], ephb2 [1370], thbs1 [1371], id3 [1372], nox4 [1373], pcsk1 [1374], wnk4 [1375], vldlr [1376], vdr [1377], gpc4 [1378], ifne [1379], zfpm2 [1380], tlr2 [1381], il6 [1382], spink1 [1383], gas6 [1384], f11r [1385], siglec15 [1386], adam12 [1387], myoc [1388], usp2 [1389], sik1 [1390], ccn2 [1391], igfbp3 [1392], lyve1 [1393], bgn [1394], eda [1395], ntrk2 [1396], slit2 [1397], irx3 [1191], nnmt [1398], chi3l1 [1399], runx1 [1400], apln [1401], pgf [1402], htr6 [742], nrg1 [1403], npy4r [1404], ccdc80 [1405], kiss1 [1406], slc6a15 [1407], esyt3 [1408], sorbs1 [1409], slc38a3 [1410], lxn [1411] and trh [1412] are potential targets for obesity. Studies have shown that genes including vcam1 [1413], stra6 [1414], aqp8 [1415], fabp4 [1416], flt1 [1417], bmp4 [633], prl [1418], adamts9 [1419], ndnf [1420], dtx1 [1421], cd4 [1422], adamts5 [1423], mmp8 [1424], adrb2 [1425], ptprd [1426], insig1 [1427], lepr [1428], il15 [1429], apoe [1430], ace [1431], lrp1b [1432], alpp [1433], hkdc1 [1434], pnpla3 [1435], fads1 [1436], mmp3 [1437], fgfr2 [1438], il33 [1439], foxc2 [1440], fst [1441], hoxc8 [1442], bdnf [1443], ndrg2 [1444], angptl4 [1445], tgfbi [1446], vdr [1447], gpc4 [1448], cyp24a1 [1449], tlr2 [1450], il6 [1451], klf2 [1452], igfbp3 [1453], slit2 [1454], apln [1455], notch2 [1456], pgf [1457], nrg1 [1458], tlr1 [1459], ccdc80 [1460] and kiss1 [1461] are the contributing factors to gestational diabetes mellitus pathogenesis. A previous study identified genes including vcam1 [1462], stra6 [1013], wnt6 [1463], fabp4 [1464], sox6 [1465], pla2g7 [1466], mmp12 [1467], faim2 [770], snca [1468], apod [1469], lag3 [1470], prex1 [1471], flt1 [1472], fabp5 [1473], trpc3 [1474], thsd7a [1475], prkcb [1476], pde3b [1477], bmp4 [1478], prl [1479], nefl [1480], wnt5a [1481], timp3 [1306], cpe [1482], adamts9 [1483], ndnf [1484], snap25 [1485], scd [1058], ldlr [1486], cd4 [1487], trpa1 [1488], rspo3 [1489], pde4b [1490], th [1491], cacna1d [1492], mmp8 [1318], kdr [1493], adrb2 [653], kl [1494], plau [1321], ccnd2 [1495], ptprd [1496], siglec1 [1497], ucn2 [1324], cyp2j2 [1498], cyp1a1 [1499], ntn1 [1090], mstn [1326], lepr [1500], il15 [1501], bmp2 [1502], apoe [1503], cubn [1504], casp1 [1505], mgll [1506], efnb2 [1507], nid1 [1508], abcg2 [1102], ace [1509], stmn2 [1510], ica1 [1511], trpc6 [1512], gpbar1 [1513], elmo1 [1514], doc2b [1515], ank1 [1516], sorcs1 [1517], hkdc1 [1518], pnpla3 [1519], fads1 [1520], acsl5 [1521], hla-c [1522], s100a4 [1523], cort [1524], mmp3 [1525], gdnf [1526], cables1 [1354], il33 [1527], ngf [1528], foxc2 [1529], comp [1530], fst [1531], sorbs2 [1532], gata6 [1533], ptgs2 [1141], dact1 [1534], dgat2 [1535], bdnf [1484], ndrg2 [1536], sgk1 [1537], tlr3 [1152], angptl4 [1538], ephb2 [1539], thbs1 [1540], mbp [1541], nox4 [1542], pi16 [1543], pcsk1 [888], tgfbi [1544], igf2bp1 [1545], wnk4 [1546], vldlr [1547], vdr [1548], gpc4 [1274], ptprn [1549], epha4 [712], sox9 [1550], mgp [1551], cyp24a1 [1552], tlr2 [1553], il6 [1554], nppb [1555], spink1 [1556], gas6 [1557], f11r [1558], foxd1 [1559], adam12 [1560], klf2 [1561], sik1 [1562], fgf7 [1563], igfbp3 [918], lyve1 [1393], eda [1395], slit2 [1564], irx3 [1565], nnmt [1398], chi3l1 [1566], runx1 [1567], apln [1568], col4a3 [1569], notch2 [1570], pdlim5 [740], nrg1 [1571], dmrt2 [1572], npy4r [1573], cd200 [1574], bst2 [1575], tfpi2 [1576], kiss1 [1406], mpp7 [1577], sorbs1 [1409], slc38a3 [1578], chn2 [1579] and trh [1580] have been implicated in diabetes mellitus pathology. A previous bioinformatics study suggested that genes including grp [1581], avpr1a [1582], ano1 [1583], ntrk1 [1584], fgf10 [1585], mmp12 [1586], snca [1587], ccr1 [1588], flt1 [1589], fabp5 [1590], trpc3 [1591], bmp4 [1592], prl [1593], wnt5a [1594], timp3 [1595], serpinb2 [1596], nlrp2 [1050], npw [1053], ereg [1597], ndnf [1598], snap25 [1056], syt1 [1599], cd4 [1600], gpr183 [1601], trpa1 [1602], pde4b [1603], th [1604], mmp8 [1605], adrb2 [1606], mgat3 [1607], plau [1608], astn2 [1609], il15 [1610], bmp2 [1611], apoe [1612], pde9a [1613], mgll [1100], efnb2 [1101], hmgcr [1614], ace [1615], syt9 [1616], trpc6 [1617], xcr1 [1618], s100a4 [1619], mmp3 [1620], gdnf [1621], il33 [1622], ngf [1623], gdf6 [1133], comp [1624], acan [1625], ptgs2 [1626], gdf10 [1627], bdnf [1628], ndrg2 [1629], sgk1 [1630], tlr3 [1631], ephb2 [1632], mbp [1633], nox4 [1619], shank2 [1634], pi16 [1635], dlx5 [1636], vdr [1637], zfhx2 [1638], epha4 [1639], cyp24a1 [1640], id2 [1641], tlr2 [1642], il6 [1643], spink1 [1644], gas6 [1645], klf2 [1646], six1 [1647], chl1 [1648], slit2 [1649], runx1 [1650], notch2 [1651], pgf [1652], nrg1 [1653], nalcn [1654], lxn [1655] and trh [1656] might play a role in the development of pain. Therefore, studying the enriched genes involved in the regulation of endometriosis might be helpful to clarify the incidence or molecular pathogenic mechanisms of various complications including ovarian cancer, endometrial cancer, cervical cancer, hypertension, cardiovascular diseases, infertility, inflammation, polycystic ovarian syndrome, obesity, gestational diabetes mellitus, diabetes mellitus and pain.

Establishing PPI network and module analysis is friendly for researchers to investigate the underlying molecular mechanism of endometriosis for the reason that the DEGs would be grouped and ordered in the network judging by their interactions. PPI network and module analyses could help to find hub genes involved in the regulation of endometriosis. A recent study suggested that the hub genes including vcam1 [110], adrb2 [126] and dapk1 [183] might take part in the progression of endometriosis. Recent evidence indicates that the hub genes including vcam1 [190], adrb2 [237], foxq1 [210], actbl2 [357], dapk1 [379], actc1 [359], cst4 [229], nfib [363], nfix [1657] and erg [1658] are potential therapeutic targets in ovarian cancer. Previous studies have reported that hub genes including vcam1 [478], foxq1 [492], dapk1 [609] and erg [1659] participate in the progression of cervical cancer. Studies have shown that hub genes including vcam1 [615], adrb2 [653] and dapk1 [751] are involved in the regulation of hypertension. Several studies have found hub genes including vcam1 [761], prkcb [780], adrb2 [808], prkd1 [920], dapk1 [947], actc1 [924] and nfix [1660] expression levels were significantly altered in cardiovascular diseases. Studies show that hub genes including vcam1 [761], snca [1026], adrb2 [1073], foxq1 [1044], prkd1 [1188], dapk1 [947], actc1 [1192], cst1 [1030], nfib [1198] and erg [1661] are mainly involved in progression of inflammation. Previous studies have shown that hub genes including vcam1 [1232] and adrb2 [969] were identified to be closely associated with polycystic ovarian syndrome. Many studies have confirmed that hub genes including vcam1 [1287], prkcb [1301] and adrb2 [1319] were an important participant in obesity. Previous studies have found that hub genes including vcam1 [1413] and adrb2 [1425] were shown to be primarily involved in gestational diabetes mellitus. A growing number of studies have demonstrated that hub genes including vcam1 [1462], snca [1468], prkcb [1476] and adrb2 [653] play an important role in progression of diabetes mellitus. Accumulating evidence shows that hub gene adrb2 [969] is an important risk factor for infertility. Recent study reported that hub gene adrb2 [1606] plays a crucial role in pain progression. A recent study showed that hub genes including dapk1 [183], cst1 [394] and nfix [1662] plays an important role in the pathogenesis of endometrial cancer. Our findings suggested mdfi, tnfrsf19 and foxl1 as potential novel diagnostic biomarkers for endometriosis. This investigation identified the possible hub genes that were highly correlated with the PPI network to find the novel biomarkers associated in the pathogenesis of endometriosis. Our ROC curve analysis showed that hub genes have diagnostic value for endometriosis.

In this investigation, the miRNA-hub gene regulatory network and TF-hub gene regulatory network of the hub genes in endometriosis were analyzed by using miRNet and NetworkAnalyst database. These analyses could help to find some miRNAs, TFs and hub genes involved in the regulation of endometriosis. Studies have shown that biomarkers including ccnd2 [246], vcam1 [190], pdgfb [221], ptch1 [263], foxq1 [210], igf2bp1 [324], actc1 [359], epb41l3 [315], dapk1 [379], hsa-mir-17-5p [1663], tcf3 [1664], rnf2 [1665], clock [1666], smarca4 [1667] and trim28 [1668] can lead to ovarian cancer. Studies reported that biomarkers including ccnd2 [511], vcam1 [478], ptch1 [523], foxq1 [492], igf2bp1 [568], epb41l3 [560], dapk1 [609], hsa-mir-17-5p [1669], tcf3 [1670] and trim28 [1671] were proposed to contribute to the development of cervical cancer. A previous study reported that biomarkers including ccnd2 [814], vcam1 [615], pdgfb [795], prkcb [780], actc1 [924], dapk1 [947], hsa-mir-17-5p [1672], hsa-mir-2110 [1673], tcf3 [1674] and smarca4 [1675] play a key role in cardiovascular diseases. Accumulated evidence has demonstrated that biomarkers including ccnd2 [1083], vcam1 [761], pdgfb [1059], ptch1 [1098], foxq1 [1044], igf2bp1 [1162], actc1 [1192], dapk1 [947], hsa-mir-2110 [1676], hsa-mir-10b-5p [1677], tcf3 [1678], nr1i2 [1679] and trim28 [1680] are associated with inflammation. Studies have shown that biomarkers including ccnd2 [1495], vcam1 [1462], ptprd [1496], prkcb [1476], igf2bp1 [1545], hsa-mir-200a-3p [1681] and hsa-mir-10b-5p [1682] were identified to be associated with diabetes mellitus. A previous study found that biomarkers including vcam1 [110], pdgfb [121], igf2bp1 [162], dapk1 [183] and hsa-mir-17-5p [1683] have been found in endometriosis. Recent studies have identified biomarkers including vcam1 [1232], hsa-mir-17-5p [1684] and hsa-mir-2110 [1685] are involved in the pathogenesis and progression of polycystic ovarian syndrome. A previous study reported that biomarkers including vcam1 [1287], pdgfb [1311], prkcb [1301], hsa-mir-17-5p [1686], hsa-mir-10b-5p [1687] and trim28 [1688] are associated with the pathogenesis and development of obesity. Many studies have shown that biomarkers including vcam1 [1413], ptprd [1426] and hsa-mir-17-5p [1689] are likely to be important in the development of gestational diabetes mellitus. Recent studies have demonstrated that biomarkers including igf2bp1 [453], dapk1 [183], ptprd [412], tcf3 [1690], smarca4 [1691] and trim28 [1692] are important in the development of endometrial cancer. Research has shown that biomarkers including ptprd [656], pdgfb [642], dapk1 [751], hsa-mir-4432 [1693], smarca4 [1694] and trim28 [1695] might be potential therapeutic targets for hypertension. Recent studies have proposed that the biomarkers including igf2bp1 [994], hsa-mir-17-5p [1696] and smarca4 [1697] serve a vital role in infertility. Study has suggested that TRIM28 [1698] might be involved in the development of pain. New biomarkers associated with diagnosis were identified in this study: st8sia4, mdfi, krt18, stx11, hsa-mir-3143, hsa-mir-6888-5p, hsa-mir-3122, hsa-mir-556-3p, hsa-mir-1229-5p, phc1, hoxc9, prdm14 and htt (huntingtin). We suggest that exercise can regulate the expression of these miRNAs, TFs and hub genes, thereby inhibiting the occurrence and development of endometriosis.

There are few limitations in our investigation. While our investigation presents promising results, several limitations should be accepted. In vivo and in vitro validation experiments for hub genes and clinical trials are required to assess the correlation linking clinical parameters and the hub genes in endometriosis pathogenesis. In this investigation, we did not account for the potential confounding effects of demographic variables. We will conduct more in-depth research in the future.

Conclusions

The current investigation identified biomarkers and pathways which might be involved in endometriosis progression through the integrated analysis of NGS dataset. These results might contribute to a better understanding of the molecular mechanisms which underlie endometriosis and provide a series of potential biomarkers. However, further experiments are required to verify the findings of the current investigations. Therefore, further experiments with additional patient cohorts are also required to confirm the results of these investigations. In vivo and in vitro investigation of gene and pathway interaction is essential to delineate the specific roles of the identified biomarkers, which might help to confirm biomarker functions and reveal the molecular mechanisms underlying endometriosis.