Identification of platinum-resistance related small GTPase binding signatures to predict the prognosis of ovarian cancer by machine learning and integrated bioinformatic analyses

Identification of platinum-resistance related small GTPase binding signatures to predict the prognosis of ovarian cancer by machine learning and integrated bioinformatic analyses | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Identification of platinum-resistance related small GTPase binding signatures to predict the prognosis of ovarian cancer by machine learning and integrated bioinformatic analyses Ya-jun Zhong, Yi-lin Zhu, Shi-qi Wang, Yuan-rong Wang, Lan-ying Bu, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4336933/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background High grade serous ovarian cancer (HGSOC) had high lethality due to its high relapse rate and acquired drug resistance. The tumor immune microenvironment (TIME) of HGSOC was heterogeneous, and mostly immune cold. We aimed to build the bridge between platinum-resistance related signatures and patient overall survival (OS). Methods The RNA sequencing data from GSE160626 was used for extraction of platinum-resistance related genes. The TCGA-OV cohort were fitted into 101 kinds of machine learning methods, and the validation cohort included GSE9899, GSE63885 and GSE26193. Numerous methods including the Cindex, receiver operating characteristic curve (ROC), univariate and multivariate Cox regression, and the decision curve analysis (DCA) were applied to detect the performances of platinum-resistance related risk score (PRRS) and a PRRS based nomogram. The single-cell RNA sequencing data and Spatial Transcriptomics data were used to determine the risky cell types correlated with our PRRS. Results Based on platinum-resistance related genes, we conducted consensus clustering and defined a platinum-resistance resembling cluster, which had significantly shorter OS. And with DEGs related to small GTPase between two clusters, we established a PRRS and a PRRS based nomogram, which had excellent performances in predicting OS of serous ovarian cancer patients. We further determined SPP1 + M2-like Macrophages were risky factors correlated with the PRRS, and determined ABCA1 and NDRG1 as the hub genes related to patient OS. Conclusion Small GTPase was a dominant feature of platinum-resistance resembling clusters. PRRS had terrific predicting value and correlated with SPP1 + M2-like Macrophages. Ovarian cancer Platinum resistance Small GTPase Predicting model Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction High grade serous ovarian cancer (HGSOC) comprises 70–80 percentage of all ovarian cancer (OC) types, which is a leading ovarian cancer type resulting in tumor specific death[ 1 ]. Due to insufficient strategies for the early detection of OC, most patients are diagnosed with late-staged and high-graded serous OC. Currently, the treatment options for newly diagnosed HGSOC remain relatively homogeneous and fixed, consisting mainly of surgery and platinum-based chemotherapy. Normally, HGSOC responds to platinum-based chemotherapy, but after recurring, which occurs more than half of all cases, OC cells will almost always acquire resistance to platinum, causing an extremely low 5-year survival in HGSOC[ 2 , 3 ]. In this sense, existing mechanisms for assessing disease severity, including tumor grading and clinical staging, may become less robust in predicting patient overall survival and implementing customized treatment plans. Though the application of immune checkpoint blockers (ICB) has brought light to cancer immunotherapy, OC patients are less likely to benefit[ 4 , 5 ]. In a broad sense, most OC is characterized as immune cold, which may affect the efficacy of ICB treatment, but more explicit evidences are needed to systematically parse the tumor immune microenvironment (TIME) of OC[ 4 ]. And a recent study shows that targeting St3gal3 to block aberrant sialylation in OC cells can alternate the TIME by converting the M2-like Macrophages to an M1-like phenotype and thus enhance the infiltration of functional CD8 + T cells, improving the effectiveness of ICB treatment[ 6 ]. This has shed light on the immunotherapy in OC, and highlights the contribution of the interaction between cancer cells and immune cells. Small GTPase plays an important role as molecular switches in cells, profoundly affecting the biological processes[ 7 , 8 ], and emerging studies point to its role in immunity. And there have been studies disclosing the correlation between small GTPase and the regulation of Macrophages[ 9 – 11 ]. Moreover, in OC, perturbation of small GTPase contributes to a more metastatic, invasive and migrative ovarian cancer cells[ 12 , 13 ]. Simultaneously, treatments targeting small GTPase have been raised[ 14 ] and great breakthrough has been made[ 15 ]. These advances emphasize the necessity and urgency to explore the landscape of small GTPase in OC. In this study, we explored the transcriptional diversity between platinum-sensitive and resistant samples. And based on these differentially expressed genes, we divided the TCGA-OV cohort into two clusters, which had resembling gene expression features. Patients in the cluster with platinum-resistant features had significantly worse overall survival. We further detected the DEGs between the two clusters, and established a platinum-resistance related risk score using small GTPase binding related genes, which had terrific value in the prediction of OS in HGSOC cohorts. Combining with clinical information, a nomogram was constructed, and had robust efficacy. We further explored that SPP1 + M2-like Macrophages had the most expression level of risky small GTPase binding related genes with single-cell RNA and spatial transcriptomics data. Moreover, NDRG1 and ABCA1 were the hub genes for SPP1 + Macrophages in affecting OC patient survival. Methods Data acquisition The RNA sequencing data included in this study were collected from the TCGA-OV cohort (n = 377), and three micro array data based on GPL570 platform were obtained from GSE9899 (n = 167)[ 16 ], GSE63885 (n = 70)[ 17 ] and GSE26193 (n = 79)[ 18 ] with clinical information. Only patients with the diagnosis of serous ovarian cancer were taken into account. The RNA sequencing data of platinum-sensitive and resistant patients were collected from GSE160626[ 19 ], with 10 platinum-sensitive and 8 platinum-resistant samples from OC patients. The single-cell RNA sequencing data were obtained from recent research containing 7 platinum-sensitive and 3 platinum-resistant samples[ 20 ]. While the spatial transcriptomics (ST) data were available in GSE211956, with 3 good response, 2 partial response and 3 poor response samples[ 21 ]. Differentially expressed gene (DEG) detection Regarding the transcripts per kilobase million (TPM) data in GSE160624, we utilized a Wilcoxon test to detect DEGs. The significant criteria were set as an absolute log fold change (FC) greater than 1 and a p-value smaller than 0.05 (see Supplementary materials Table 1). For raw counts data in the TCGA-OV cohort, three algorithms including Limma [ 22 ], edgeR [ 23 ] and DESeq2 [ 24 ] were applied, the initial filtering criteria were set as an absolute logFC greater than 1 and a p-value smaller than 0.05, and the intersecting genes of three algorithms were considered as significant DEGs. Consensus clustering The consensus clustering was conducted with cluster algorithm set as k-means , using 26 platinum-resistant related genes (genes with a corresponding Ensembl ID, not mitochondrial genes) based on the ConsensusClusterPlus [ 25 ] R packages. The optimal k value was determined by the minimal proportion of ambiguous clustering (PCA) value. Gene enrichment analysis The Gene Ontology (GO) analysis was conducted with enrichGO command in R packages clusterProfiler [ 26 ]. And the gene set enrichment analysis (GSEA) was applied with GSEA command with the HALLMARK and KEGG gene sets. Survival analysis The Kaplan-Meier (KM) analysis was conducted with R packages survival , and the optimal cut-off value was determined using R packages survminer . A p-value smaller than 0.05 was considered as statistically significant. The KM analysis results of ABCA1 and NDRG1 were obtained in KM-plotter website ( https://kmplot.com/analysis/ )[ 27 ]. Integrated machine learning method A total of 101 machine learning methods including Random Survival Forest (RSF), Enet, StepCox, CoxBoost, plsRCox, SuperPC, GBM, survival SVM, Least Absolute Shrinkage and Selection Operator (LASSO) and Ridge, and their combinations were applied in this study. Specifically, when combining two algorithms, the first algorithm was used to select variables accordingly and the second algorithm was utilized to build the predicting model with filtered variables. For RSF, the optimal mtry value was determined by method of exhaustion. And the relative importance of variables was obtained with subsample command in R package randomForestSRC . Evaluation of immune signatures A total of 5 immune algorithms including CIBERSORT, MCPcounter, EPIC, TIMER and quantiseq were applied to evaluate the immune cell infiltration landscape in the TCGA-OV cohort using R package IOBR [ 28 ]. The immune signatures including stimulatory, inhibitory and antigen presenting molecules were summarized from previous research. Processing of the single-cell RNA sequencing data The processed data were obtained from recent research[ 20 ], and to better integrate data from different samples, we performed the reciprocal PCA (RPCA) integration (Figure S1 ). The signature score was calculated with AddModuleScore command in R package Seurat [ 29 ]. For the detection of DEGs, the criteria were set as a min.pct greater than 0.2 and a logFC greater than 0.2. Cell-cell communication analysis The intra-cellular communication patterns were detected using R package CellChat [ 30 ]. Processing of the ST data Spots in the ST data were firstly filtered with number of features greater than 300 and percentage of mitochondrial genes smaller than 0.3 as criteria. Then the normalized RNA value was obtained by the SCTransform command. The signature score was calculated using AddModuleScore command. Statistical analysis All data processing, statistical analysis, and plotting were performed in R 4.3.1 software. Correlations between two continuous variables were calculated via Sperman’s correlation analysis. The chi-squared test was used to compare categorical variables. The univariate and multivariate Cox regression analyses were conducted with R package survival . The receiver operating characteristic curve (ROC) was implemented using R package timeROC , and the time-dependent ROC were calculated with R package reportROC with a 1-month interval. All statistical tests performed were two-sided. A p-value smaller than 0.05 was considered as statistically significant. Results Consensus clustering with platinum-resistance related gene signatures defined two clusters in the TCGA-OV cohort From a previous RNA sequencing data on platinum-sensitive and resistant samples, we defined a total of 50 significant platinum-resistance related gene signatures (Fig. 1 A and Supplementary materials Table 1). After filtering mitochondrial genes and mapping to the TCGA-OV cohort, 26 genes were finally selected for the consensus clustering. Based on the PAC results, k = 2 was considered optimal with least PAC (Fig. 1 B). And the clustering results were shown in Fig. 1 C. The summarized clinical status and specified expression patterns of two clusters were shown in Fig. 1 D, where cluster 2 showed obvious elevation of platinum-resistance related genes. A total number of 4218 differentially expressed genes (DEGs) was detected between C2 and C1 (see Supplementary materials Table 2–5). Using a GO enrichment analysis, the DEGs between two clusters were specifically enriched in small GTPase and GTPase activity and in the molecular function related to GTPase binding and small GTPase binding (Fig. 1 E). Based on the KM results, there was significant survival advantages in the cluster 1 (Fig. 1 F), while there were no significant differences in the distribution of tumor grading and clinical staging between two clusters (Fig. 1 G), as most patients suffered from high-graded and late-staged ovarian cancer. Moreover, proliferating related signatures TUBA4A and PROM2 were significantly higher in the cluster 2, and PROM2 was also proved to correlate with cancer stem cells (Fig. 1 H). Establishment of the platinum-resistance related risk score (PRRS) and nomogram with machine learning methods A total of 108 DEGs from two clusters within the GO small GTPase binding molecular function gene list were selected to build the PRRS. As was shown in Fig. 2 A, in the TCGA-OV cohort, 101 kinds of prediction models were fitted and the C-index was calculated among 4 cohorts for validation. The RSF method was considered optimal with highest mean C-index. Figure 2 B illustrated the top 20 important variables based on the RSF results. Based on the PRRS, in the combinate META-OV cohort and a single TCGA-OV cohort, with medium cut-off, patients with higher PRRS had significantly worse overall survival (OS) (Fig. 2 C, 2 D). And with optimal cut-off, PRRS in all the validation cohorts significantly represented worse survival (Fig. 2 E- 2 G). To test the efficacy of PRRS, we conducted univariate and multivariate Cox regression analyses. In the univariate Cox regression results, RiskScore, together with Grade 3/4 when compared with G1/2, Stage III/IV when compared with Stage I/II exhibited a Hazard ration (HR) greater than one, meaning they were risky factors for ovarian cancer (Fig. 3 A). And in the multivariable Cox regression results, the RiskScore and Stage III/IV remained HR greater than one, indicating their robust predicting value for clinical outcomes (Fig. 3 B). In order to build the bridge between the PRRS and clinical indicators, a nomogram was established (Fig. 3 C). And the calibration curves of 1-year, 3-year and 5-year survival demonstrated that the nomogram had excellent predicting accuracy (Fig. 3 D). Moreover, the area under curve (AUC) value was 0.83 for 1 year, 0.848 for 3 years, and 0.838 for 5 years, indicating the terrific performance of the nomogram (Fig. 3 E). We also conducted the time dependent AUC changes at 1 month interval, and discovered that the nomogram and PRRS had great advantages as compared with Grade and Stage to predict the patient OS, and the nomogram was slightly more superb in predicting long time OS when compared with a single PRRS (Fig. 3 F). And the decision curve analysis (DCA) results also supported the above conclusions (Fig. 3 G). These results jointed demonstrated that PRRS and a PRRS based nomogram had great clinical significances. The immune landscape and mutation landscape between two risk subgroups and their biological pathway differences To comprehensively evaluate the immune landscape between two risk subgroups, 5 immune algorithms including CIBERSORT, MCPcounter, EPIC, TIMER and quantiseq were conducted based on the TCGA cohort. The summarized heatmap was shown in Fig. 4 A. And judging from a holistic point of view, myeloid lineages including Macrophages M2 in CIBERSORT and quantiseq, Monocytic lineage in MCPcounter, and DC in timer were significantly higher in the risk high subgroup, and neutrophils in MCPcounter and TIMER represented higher infiltration levels in the risk high subgroup. And the spearman correlation results proved evidence for the positive coupling of PRRS with above immune cells (Fig. 4 B- 4 G). Regarding stimulatory molecules, CD40LG was higher in the risk low subgroup, while ENTPD1 and ICOSLG had higher RNA level in the risk high subgroup. Inhibitory molecules, including ARG1, CD276, TGFB1 and VSIR had significant higher expression in the risk high subgroup. And there were no significant differences in antigen presenting molecules among two subgroups (Fig. 4 H). Then we explored the biological pathway differences using DEGs from the Limma results. The gene set enrichment analysis (GSEA) results of HALLMARK gene set revealed that multiple tumor progression related hallmarks, including the MYC targets, PI3K AKT MTOR signaling, together with metabolic hallmarks, such as the Adipogenesis, glycolysis and fatty acid metabolism, were significantly elevated in the risk high subgroup, which indicated the more aggressive and proliferative phenotypes in the risk high subgroup (Fig. 5 A). And the GO results also validated that the DEGs were enriched in metabolic pathways and pathways in cancer (Fig. 5 B). Moreover, we explored the mutation landscape between two subgroups, while TP53 and TTN were the most common mutation in both groups, other mutations had different distributions. For example, the risk high subgroup had more CSMD3, MUC16 and USH2A mutations, while the risk low subgroup had more AHNAK, FAT3 and HUWE1 mutations (Fig. 5 C, 5 D). Exploring the PRRS signatures with single-cell RNA sequencing data To further determined how PRRS signatures affected the tumor microenvironment (TME), a total of 10 single-cell RNA data from patients with high-grade serous OC (HGSOC) was obtained from recent research[ 20 ], and the number of filtered cells and patients’ response to chemotherapy were shown in Fig. 6 A (see Supplementary materials Table 6). A total of 9 major cell types were determined (Fig. 6 B) with canonical cell markers (Cancer cells: KRT18, CD24; Immune cells: PTPRC; T cells: CD3E; NK cells: GNLY; B/Plasma cells: CD79A; Myeloid cells: LYZ; Macrophages: C1QA; DC: C1DC; Endothelial cells: VWF; Fibroblast: COL1A2, DCN) (Figure S1 ). And the top 5 markers for each major cell type were shown in Fig. 6 C. We calculated the ratio of observed to expected cell numbers ( Ro/e ) between resistant and sensitive patients, and consistent with the cell deconvolution results, the Macrophages were more enriched in the resistant samples (Fig. 6 D). We then calculated the PRRS signature scores, and discovered that the cells in the resistant samples compared with the sensitive samples, and especially high in the Macrophages in the resistant samples (Fig. 6 E), meaning that PRRS signature genes most correlated with Macrophages. We further divided the Macrophages into 7 subgroups (Fig. 6 F), and the comparing results indicated the subgroup Macro4 in the resistant subgroup had highest PRRS signature score among other subgroups (Fig. 6 G). Compared with Macro4 sensitive cells, SPP1 had remarkably higher expression in the Marco4 resistant cells, which was known markers for M2 Macrophages (Fig. 7 A). These elevated genes were enriched in inflammatory response and angiogenesis, providing evidence for their pro-tumor role (Fig. 7 B). Using the cell chat, we further explored the outgoing signaling patterns and incoming signaling patterns (Fig. 7 C, 7 D), illustrating the special communication patterns in Macro4 resistant cells. Figure 7 E showed the significant ligand-receptor pairs, and Macro4 resistant cells had the strongest communication probability of SPP1 signaling pathways among other cells. And Fig. 7 F illustrated the specific ligands and receptors among cell types. NDRG1 and ABCA1 were the hub genes for SPP1 + Macrophages in affecting OC patient survival Among these modeled genes of PRRS, NDRG1 and ABCA1 were also markers for Macro4 resistant cells (Fig. 8 A). A higher level of ABCA1 and NDRG1 (DRG1) represented worse survival (Fig. 8 B- 8 E). Based on ST results, we explored the distribution patterns and gene expression patterns of samples from HGSOC patients who received taxane- and platinum-based neoadjuvant chemotherapy. In the poor response group, there were specific area co-expressing ABCA1, NDRG1 and had especially higher Macro4 resistant signature score, specifically in P1 and P8 (Fig. 8 F). While in the good response and partial response group, the co-expressing patterns was low and insufficient (Fig. 8 G- 8 H). These results also proved evidence for the high RNA level of ABCA1 and NDRG1 in OC, paving way for their significant role in OC. Discussion The acquired resistance to chemotherapy has always been a hot topic in the management of OC patients. With its high reoccurrence rate and severely high lethality, resistant OC after relapse normally results in a poor overall survival[ 31 ]. The current assessment systems for serous OC failed to distinguish high risk patients, as most of them were diagnosed with advanced HGSOC. Though current research managed to build a scoring system for OC[ 32 , 33 ] and explore the risky cell types[ 34 ], and some attempted to build the bridge between drug resistance and patient survival[ 33 , 35 ], a more efficient model was of necessity. In our study, we identified platinum-resistant gene signatures, and the clustering results indicated a platinum-resistance resembling cluster, which had significantly worse OS. Interestingly, these DEGs in the platinum-resistance resembling cluster were enriched in small GTPase related gene sets. As a result, via integrated machine learning methods, we built a PRRS model based on these small GTPase related DEGs, which had ideal and stable performances among validation cohorts. Combined with current clinical indexes, a nomogram was produced and exhibited terrific predicting performances in predicting patient survival, which had a settled AUC value greater than 0.8, meaning high accuracy and sensitivity. With the advances in sequencing technology and the wide application of single-cell RNA sequencing and ST technology, emerging studies attempted to parse the OC in a more nuanced perspective and highlighted the significances of TIME[ 20 , 21 ]. Using deconvolution methods, we discovered a robust correlation with PRRS and M2 Macrophages. And the results were validated by the single-cell RNA sequencing data, where SPP1 + M2-like Macrophages had the most expression of small GTPase related signatures, indicating a vital role of small GTPase molecules in regulating M2-like Macrophages. Via cell chat, we discovered that the specific SPP1 + Macrophages had unique intracellular communication patterns, especially their dominant contributions in the SPP1 signaling pathway. In the last section of our study, we identified ABCA1 and NDRG1 (DRG1) as the hub genes for SPP1 + Macrophages in affecting patient OS. Based on ST results, we discovered the co-localization of SPP1 + Macrophages markers with ABCA1 and NDRG1. A previous research has also discovered ABCA1 mediates the acquired resistance in ovarian cancer cells, and correlates with poor OS[ 36 ]. Acting as an important myeloid leukocyte activator, the role of NDRG1 in cancer was critical, and in ovarian cancer, knockdown of NDRG1 results in enhanced tumor cell adhesion, migration and invasion activities[ 37 , 38 ]. But the role of ABCA1 and NDRG1 in regulating Macrophages in OC remained unclear. The present research showed the important role of ABCA1 and NDRG1 in SPP1 + Macrophages, and deserved future research. In conclusion, our study identified a cluster of patients with platinum-resistant features, which had comparable shorter OS. We discovered that DEGs in the platinum-resistance resembling clusters were correlated with small GTPase. A PRRS model was built accordingly and was competent in detecting high risk patients among cohorts. Moreover, we firstly discussed the possible significance of ABCA1 and NDRG1 in SPP1 + Macrophages, providing new perspectives for future research. This work had its limitations, the PRRS model and nomogram needed to be fully optimized with more clinical sample. And the exact mechanisms of ABCA1 and NDRG1 in regulating Macrophages required more in vitro and in vivo experiments. Abbreviations HGSOC : High grade serous ovarian cancer TIME : Tumor immune microenvironment OS : Overall survival ROC : Receiver operating characteristic curve DCA : Decision curve analysis PRRS : Platinum-resistance related risk score ICB : Immune checkpoint blockers ST : Spatial transcriptomics DEG : Differentially expressed gene TPM : Transcripts per kilobase million FC : Fold change GSEA : Gene set enrichment analysis KM : Kaplan-Meier RSF : Random Survival Forest LASSO : Least Absolute Shrinkage and Selection Operator RPCA : Reciprocal PCA PAC : Proportion of ambiguous clustering GO : Gene Ontology BP : Biological process MF : molecular function Ro/e : Ratio of observed to expected cell numbers Declarations Availability of data and materials All data included in this study is available in the TCGA and GEO databases. All codes used for analyses are available on reasonable request Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests None. Acknowledgements Not applicable. Funding Not applicable. Contributions HZ, WK and YZ proposed and designed this study. YZ, YW and WK conducted most of the formal analyses. YZ and SW made contributions to the embellishment of figures and wrote this manuscript. LB and RZ collected the data used for analyses. YZ and YW reviewed the data. All authors confirmed the final version for submission. References Bowtell DD, Böhm S, Ahmed AA, Aspuria PJ, Bast RC Jr., Beral V, Berek JS, Birrer MJ, Blagden S, Bookman MA, et al. Rethinking ovarian cancer II: reducing mortality from high-grade serous ovarian cancer. Nat Rev Cancer. 2015;15(11):668–79. Matulonis UA, Sood AK, Fallowfield L, Howitt BE, Sehouli J, Karlan BY. Ovarian cancer. Nat Reviews Disease Primers. 2016;2(1):16061. Wang L, Wang X, Zhu X, Zhong L, Jiang Q, Wang Y, Tang Q, Li Q, Zhang C, Wang H, et al. Drug resistance in ovarian cancer: from mechanism to clinical trial. Mol Cancer. 2024;23(1):66. Pawłowska A, Rekowska A, Kuryło W, Pańczyszyn A, Kotarski J, Wertel I. Current Understanding on Why Ovarian Cancer Is Resistant to Immune Checkpoint Inhibitors. Int J Mol Sci 2023, 24(13). Wang ZB, Zhang X, Fang C, Liu XT, Liao QJ, Wu N, Wang J. Immunotherapy and the ovarian cancer microenvironment: Exploring potential strategies for enhanced treatment efficacy. Immunology; 2024. Cao K, Zhang G, Yang M, Wang Y, He M, Zhang C, Huang Y, Lu J, Liu H. Attenuation of Sialylation Augments Antitumor Immunity and Improves Response to Immunotherapy in Ovarian Cancer. Cancer Res. 2023;83(13):2171–86. Johnson DS, Chen YH. Ras family of small GTPases in immunity and inflammation. Curr Opin Pharmacol. 2012;12(4):458–63. Song S, Cong W, Zhou S, Shi Y, Dai W, Zhang H, Wang X, He B, Zhang Q. Small GTPases: Structure, biological function and its interaction with nanoparticles. Asian J Pharm Sci. 2019;14(1):30–9. Cai Y, Sukhova GK, Wong HK, Xu A, Tergaonkar V, Vanhoutte PM, Tang EH. Rap1 induces cytokine production in pro-inflammatory macrophages through NFκB signaling and is highly expressed in human atherosclerotic lesions. Cell cycle (Georgetown Tex). 2015;14(22):3580–92. Carvalho BC, Oliveira LC, Rocha CD, Fernandes HB, Oliveira IM, Leão FB, Valverde TM, Rego IMG, Ghosh S, Silva AM. Both knock-down and overexpression of Rap2a small GTPase in macrophages result in impairment of NF-κB activity and inflammatory gene expression. Mol Immunol. 2019;109:27–37. Tang S, Chen T, Yang M, Wang L, Yu Z, Xie B, Qian C, Xu S, Li N, Cao X, et al. Extracellular calcium elicits feedforward regulation of the Toll-like receptor-triggered innate immune response. Cell Mol Immunol. 2017;14(2):180–91. Gao L, Zheng M, Guo Q, Nie X, Li X, Hao Y, Liu J, Zhu L, Lin B. Downregulation of Rab23 inhibits proliferation, invasion, and metastasis of human ovarian cancer. Int J Biochem Cell Biol. 2019;116:105617. Guo J, Han X, Li J, Li Z, Yi J, Gao Y, Zhao X, Yue W. Single-cell transcriptomics in ovarian cancer identify a metastasis-associated cell cluster overexpressed RAB13. J translational Med. 2023;21(1):254. Prieto-Dominguez N, Parnell C, Teng Y. Drugging the Small GTPase Pathways in Cancer Treatment: Promises and Challenges. Cells 2019, 8(3). Yin G, Huang J, Petela J, Jiang H, Zhang Y, Gong S, Wu J, Liu B, Shi J, Gao Y. Targeting small GTPases: emerging grasps on previously untamable targets, pioneered by KRAS. Signal Transduct Target Therapy. 2023;8(1):212. Tothill RW, Tinker AV, George J, Brown R, Fox SB, Lade S, Johnson DS, Trivett MK, Etemadmoghadam D, Locandro B, et al. Novel molecular subtypes of serous and endometrioid ovarian cancer linked to clinical outcome. Clin cancer research: official J Am Association Cancer Res. 2008;14(16):5198–208. Lisowska KM, Olbryt M, Dudaladava V, Pamuła-Piłat J, Kujawa K, Grzybowska E, Jarząb M, Student S, Rzepecka IK, Jarząb B, et al. Gene expression analysis in ovarian cancer - faults and hints from DNA microarray study. Front Oncol. 2014;4:6. Mateescu B, Batista L, Cardon M, Gruosso T, de Feraudy Y, Mariani O, Nicolas A, Meyniel JP, Cottu P, Sastre-Garau X, et al. miR-141 and miR-200a act on ovarian tumorigenesis by controlling oxidative stress response. Nat Med. 2011;17(12):1627–35. Wu Y, Xu M, Feng Z, Wu H, Wu J, Ha X, Wu Y, Chen S, Xu F, Wen H, et al. AUF1-induced circular RNA hsa_circ_0010467 promotes platinum resistance of ovarian cancer through miR-637/LIF/STAT3 axis. Cell Mol Life Sci. 2023;80(9):256. Zheng X, Wang X, Cheng X, Liu Z, Yin Y, Li X, Huang Z, Wang Z, Guo W, Ginhoux F, et al. Single-cell analyses implicate ascites in remodeling the ecosystems of primary and metastatic tumors in ovarian cancer. Nat cancer. 2023;4(8):1138–56. Denisenko E, de Kock L, Tan A, Beasley AB, Beilin M, Jones ME, Hou R, Muirí D, Bilic S, Mohan G, et al. Spatial transcriptomics reveals discrete tumour microenvironments and autocrine loops within ovarian cancer subclones. Nat Commun. 2024;15(1):2860. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47. Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinf (Oxford England). 2010;26(1):139–40. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinf (Oxford England). 2010;26(12):1572–3. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7. Győrffy B. Transcriptome-level discovery of survival-associated biomarkers and therapy targets in non-small-cell lung cancer. Br J Pharmacol. 2024;181(3):362–74. Zeng D, Ye Z, Shen R, Yu G, Wu J, Xiong Y, Zhou R, Qiu W, Huang N, Sun L, et al. IOBR: Multi-Omics Immuno-Oncology Biological Research to Decode Tumor Microenvironment and Signatures. Front Immunol. 2021;12:687975. Hao Y, Hao S, Andersen-Nissen E, Mauck WM 3rd, Zheng S, Butler A, Lee MJ, Wilk AJ, Darby C, Zager M, et al. Integrated analysis of multimodal single-cell data. Cell. 2021;184(13):3573–e35873529. Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12(1):1088. Marchetti C, De Felice F, Romito A, Iacobelli V, Sassu CM, Corrado G, Ricci C, Scambia G, Fagotti A. Chemotherapy resistance in epithelial ovarian cancer: Mechanisms and emerging treatments. Sem Cancer Biol. 2021;77:144–66. Yang J, Wang C, Zhang Y, Cheng S, Wu M, Gu S, Xu S, Wu Y, Wang Y. A novel autophagy-related gene signature associated with prognosis and immune microenvironment in ovarian cancer. J ovarian Res. 2023;16(1):86. Wang J, Zhang X, Zheng F, Yang Q, Bi F. Mitophagy-related long non-coding RNA signature predicts prognosis and drug response in Ovarian Cancer. J ovarian Res. 2023;16(1):177. Olbrecht S, Busschaert P, Qian J, Vanderstichele A, Loverix L, Van Gorp T, Van Nieuwenhuysen E, Han S, Van den Broeck A, Coosemans A, et al. High-grade serous tubo-ovarian cancer refined with single-cell RNA sequencing: specific cell subtypes influence survival and determine molecular subtype classification. Genome Med. 2021;13(1):111. Chen L, Gao W, Lin L, Sha C, Li T, Chen Q, Wei H, Yang M, Xing J, Zhang M, et al. A methylation- and immune-related lncRNA signature to predict ovarian cancer outcome and uncover mechanisms of chemoresistance. J ovarian Res. 2023;16(1):186. Wang W, Lokman NA, Noye TM, Macpherson AM, Oehler MK, Ricciardelli C. ABCA1 is associated with the development of acquired chemotherapy resistance and predicts poor ovarian cancer outcome. Cancer drug Resist (Alhambra Calif). 2021;4(2):485–502. Zhao G, Chen J, Deng Y, Gao F, Zhu J, Feng Z, Lv X, Zhao Z. Identification of NDRG1-regulated genes associated with invasive potential in cervical and ovarian cancer cells. Biochem Biophys Res Commun. 2011;408(1):154–9. Wang B, Li J, Ye Z, Li Z, Wu X. N-myc downstream regulated gene 1 acts as a tumor suppressor in ovarian cancer. Oncol Rep. 2014;31(5):2279–85. Additional Declarations No competing interests reported. Supplementary Files FigS1.png Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4336933","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":298917766,"identity":"54072eba-d688-4d12-ab5d-10a4767b2a83","order_by":0,"name":"Ya-jun Zhong","email":"","orcid":"","institution":"Suzhou Ninth Hospital Affiliated to Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Ya-jun","middleName":"","lastName":"Zhong","suffix":""},{"id":298917772,"identity":"9edbddb6-560c-423f-a7db-77a868f52d5a","order_by":1,"name":"Yi-lin Zhu","email":"","orcid":"","institution":"Clinical Medical College of Yangzhou University","correspondingAuthor":false,"prefix":"","firstName":"Yi-lin","middleName":"","lastName":"Zhu","suffix":""},{"id":298917777,"identity":"f09d634b-e328-4dac-9c21-a31df941c82e","order_by":2,"name":"Shi-qi Wang","email":"","orcid":"","institution":"Kunming University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Shi-qi","middleName":"","lastName":"Wang","suffix":""},{"id":298917782,"identity":"5d85adbb-408a-4182-b5f2-0ad043327cac","order_by":3,"name":"Yuan-rong Wang","email":"","orcid":"","institution":"Nanjing Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yuan-rong","middleName":"","lastName":"Wang","suffix":""},{"id":298917784,"identity":"970fca29-fc9e-4852-9b2b-55335a26b0fd","order_by":4,"name":"Lan-ying Bu","email":"","orcid":"","institution":"Suzhou Ninth Hospital Affiliated to Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Lan-ying","middleName":"","lastName":"Bu","suffix":""},{"id":298917786,"identity":"fe190d21-46fe-4596-9c30-e3c4e78f3419","order_by":5,"name":"Rui-heng Zhao","email":"","orcid":"","institution":"Suzhou Ninth Hospital Affiliated to Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Rui-heng","middleName":"","lastName":"Zhao","suffix":""},{"id":298917787,"identity":"6322ef1a-8046-494d-8589-e7f833448523","order_by":6,"name":"Ying Zhou","email":"","orcid":"","institution":"Suzhou Ninth Hospital Affiliated to Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Ying","middleName":"","lastName":"Zhou","suffix":""},{"id":298917788,"identity":"defc127e-0412-4d83-a911-31e954d97988","order_by":7,"name":"Wei-yu Kong","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCElEQVRIiWNgGAWjYBACCTjF3gBmMYNJHqK08BxgbCBFC4iVwNgA5+DTItnee/g1T4VF4oabz58/5mG4w647I4Hxwds2BnlzHFqkec6lWfOckUjccDvHsJmH4Rmz2Y0EZsO5bQyGOxuwa5GTyDEzzm0Da2EEajkM0sImzdvGkGBwAJ+Wf0AtN48/hGlh/41Pi7REjvHj3AaglhsMhnBbmPFpkew5Y8b855iE8cwzOYYz5xgAtZx52Cw555yE4QYcWiSO9xh/nFFTJ9t3/PiDD28qDiebHU8++OFNmY08LluAgA0UN44NYLYBQzIDAzh+JHCqBwLmD0DCHsazw6d0FIyCUTAKRiYAANXfWwoP8ihVAAAAAElFTkSuQmCC","orcid":"","institution":"First Clinical Medical College of Nanjing Medical University","correspondingAuthor":true,"prefix":"","firstName":"Wei-yu","middleName":"","lastName":"Kong","suffix":""},{"id":298917789,"identity":"6c5c4292-c1f8-4a97-aec2-101e857b5139","order_by":8,"name":"Hong Zhou","email":"","orcid":"","institution":"Suzhou Ninth Hospital Affiliated to Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Hong","middleName":"","lastName":"Zhou","suffix":""}],"badges":[],"createdAt":"2024-04-28 08:21:56","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4336933/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4336933/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":56037508,"identity":"a75e51df-f3d6-4d37-8c0b-0a8909216b34","added_by":"auto","created_at":"2024-05-07 18:51:06","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":976979,"visible":true,"origin":"","legend":"\u003cp\u003eConsensus clustering based on platinum-resistance related gene signatures.\u003c/p\u003e\n\u003cp\u003e(A) Identification of the platinum-resistance related genes using a Wilcoxon test. (B) The proportion of ambiguous clustering (PAC) score, suggesting optimal k (k=2) by the lowest PAC. (C) The consensus clustering results with k=2. (D) The expression patterns of platinum-resistance related genes in the TCGA-OV cohort. (E) The results of GO biological process (BP) and molecular function (MF) results. (F) KM plot showed significant survival differences between two clusters. (G) Bar plot showed the proportion of different tumor grading and clinical staging in two clusters. (H) The expression of TUBA4A and PROM2 in two clusters. (***, p\u0026lt;0.001)\u003c/p\u003e","description":"","filename":"Fig1.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/7baa32dec23205b332f63e92.png"},{"id":56037503,"identity":"7de6786e-ce87-4ff9-932d-da504e5ab924","added_by":"auto","created_at":"2024-05-07 18:51:04","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1208225,"visible":true,"origin":"","legend":"\u003cp\u003eEstablishment and validation of the platinum-resistance related risk score (PRRS) via the machine learning-based integrative procedure.\u003c/p\u003e\n\u003cp\u003e(A) A total of 101 kinds of prediction models and the Cindex of each model. (B) The top 20 important variables calculated by RSF model. (C-G) KM curves of OS according to the PRRS in META-OV (p\u0026lt;0.001, medium cut-off) (C), TCGA-OV (p\u0026lt;0.001, medium cut-off) (D), GSE63885 (p=0.003, optimal cut-off) (E), GSE9899 (p=0.01, optimal cut-off) (F), GSE26193 (p=0.037, optimal cut-off) (G).\u003c/p\u003e","description":"","filename":"Fig2.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/feaabca372fd72dceb357697.png"},{"id":56037506,"identity":"33ffb46b-8101-4071-8f13-c4718f5e3779","added_by":"auto","created_at":"2024-05-07 18:51:05","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":425061,"visible":true,"origin":"","legend":"\u003cp\u003eAssessment of the PRRS and establishment of the nomogram.\u003c/p\u003e\n\u003cp\u003e(A-B) The univariable (A) and multivariable (B) Cox regression analysis results of PRRS and clinical variables. (C) Establishment of a survival nomogram. (D) A calibration plot showed the performances of the nomogram in predicting 1-year, 3-year and 5-year survival. (E) ROC curves of the nomogram to predict the patients in the Meta-OV cohort at 1, 3 and 5 years. (F) Comparison of the ROC value showed the nomogram had best predicting value. (G) The decision curve analysis (DCA) showed moderate benefit in the nomogram.\u003c/p\u003e","description":"","filename":"Fig3.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/947c026aed594ea1db512a00.png"},{"id":56037509,"identity":"042b6f4f-3984-4ff7-b39e-d5d74c5238f7","added_by":"auto","created_at":"2024-05-07 18:51:06","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1086585,"visible":true,"origin":"","legend":"\u003cp\u003eThe correlation between PRRS subgroups and the immune landscape.\u003c/p\u003e\n\u003cp\u003e(A) Heatmap showed the estimated infiltrated immune cells with 5 immune algorithms. (*, p\u0026lt;0.05; **, p\u0026lt;0.01; ***, p\u0026lt;0.001) (B-G) The Spearman correlation between PRRS and immune signatures. (H) The immune-related molecules compared between two PRRS subgroups.\u003c/p\u003e","description":"","filename":"Fig4.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/2faa89e372ae89726765d479.png"},{"id":56037505,"identity":"f79551f7-da6e-4c65-ba31-2c1850ecf28d","added_by":"auto","created_at":"2024-05-07 18:51:05","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":587417,"visible":true,"origin":"","legend":"\u003cp\u003eThe differences in biological pathways and mutational landscapes between two PRRS subgroups.\u003c/p\u003e\n\u003cp\u003e(A-B) Enrichment of differentially expressed genes between two PRRS subgroups with Hallmark (A) and GO BP (B) gene sets. (C-D) The most mutated genes in risk high (C) and risk low (D) subgroups.\u003c/p\u003e","description":"","filename":"Fig5.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/f734f7cc32cf4886397a0e9c.png"},{"id":56038229,"identity":"bc1cec48-e63a-4c05-831f-fbec54d85e67","added_by":"auto","created_at":"2024-05-07 18:59:07","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":905676,"visible":true,"origin":"","legend":"\u003cp\u003eSingle-cell RNA analysis identified a subgroup of macrophages with high small GTPase binding signatures.\u003c/p\u003e\n\u003cp\u003e(A) Composition of patients in the scRNA cohort. (B) Annotated cell types in the resistant and sensitive scRNA samples. (C) The top five markers for each cell type. (D) Tissue preference of each major cluster in Resistant and sensitive samples estimated by Ro/e. (E) The risk score modeled gene signature scores between resistant and sensitive samples and among cell types. (F) Umap plot showed 7 sub-clusters of Macrophages. (G) The modeled gene signature scores among Macrophage sub-clusters.\u003c/p\u003e","description":"","filename":"Fig6.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/fd1d36c2298aeb8f52391da7.png"},{"id":56037507,"identity":"19399e16-1371-40f3-8480-7e5a7d1cfac9","added_by":"auto","created_at":"2024-05-07 18:51:05","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":931912,"visible":true,"origin":"","legend":"\u003cp\u003eGene expression patterns and cell communication characteristics of specific group of Macrophages.\u003c/p\u003e\n\u003cp\u003e(A) The differentially expressed genes in the Macro4_Resistant sub-cluster. (B) The GO BP and MF results of differentially expressed genes. (C-D) The outgoing signaling (C) and incoming signaling (D) patterns among cell types. (E) The most important communicating pathways between Macrophages and cancer cells. (F) The SPP1 signaling pathway signatures among cell types.\u003c/p\u003e","description":"","filename":"Fig7.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/8fd0c5d3b53cbbd1f0aaba70.png"},{"id":56037510,"identity":"63dbb47e-311e-4b60-8bdf-45e91867a205","added_by":"auto","created_at":"2024-05-07 18:51:07","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":5531977,"visible":true,"origin":"","legend":"\u003cp\u003eABCA1 and NDRG1 are the hub modeled genes in SPP1 positive Macrophages in affecting patient survival.\u003c/p\u003e\n\u003cp\u003e(A) Venn plot showed the intersecting genes ABCA1 and NDRG1. (B-C) KM plot showed the OS and PFS differences between ABCA1 high and low subgroups. (D-E) KM plot showed the OS and PFS differences between NDRG1 (DRG1) high and low subgroups. (F-H) Spatial transcriptomics results of Macro4_resistant marker score, ABCA1 and NDRG1 in poor response (F), good response (G) and partial response (H) samples.\u003c/p\u003e","description":"","filename":"Fig8.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/af0e0081435109a0ed4e025f.png"},{"id":66637961,"identity":"4a35bdb6-3098-4a73-8ef2-93b1d225402c","added_by":"auto","created_at":"2024-10-15 05:55:50","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9292432,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/4c10d78c-9f10-44c3-976d-8e7c2eadff1f.pdf"},{"id":56037501,"identity":"9856e052-2b1b-41d2-af77-34a3a5a97703","added_by":"auto","created_at":"2024-05-07 18:51:03","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":897423,"visible":true,"origin":"","legend":"","description":"","filename":"FigS1.png","url":"https://assets-eu.researchsquare.com/files/rs-4336933/v1/4737cd4629a3c4da4861bca2.png"}],"financialInterests":"No competing interests reported.","formattedTitle":"Identification of platinum-resistance related small GTPase binding signatures to predict the prognosis of ovarian cancer by machine learning and integrated bioinformatic analyses","fulltext":[{"header":"Introduction","content":"\u003cp\u003eHigh grade serous ovarian cancer (HGSOC) comprises 70\u0026ndash;80 percentage of all ovarian cancer (OC) types, which is a leading ovarian cancer type resulting in tumor specific death[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Due to insufficient strategies for the early detection of OC, most patients are diagnosed with late-staged and high-graded serous OC. Currently, the treatment options for newly diagnosed HGSOC remain relatively homogeneous and fixed, consisting mainly of surgery and platinum-based chemotherapy. Normally, HGSOC responds to platinum-based chemotherapy, but after recurring, which occurs more than half of all cases, OC cells will almost always acquire resistance to platinum, causing an extremely low 5-year survival in HGSOC[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In this sense, existing mechanisms for assessing disease severity, including tumor grading and clinical staging, may become less robust in predicting patient overall survival and implementing customized treatment plans.\u003c/p\u003e \u003cp\u003eThough the application of immune checkpoint blockers (ICB) has brought light to cancer immunotherapy, OC patients are less likely to benefit[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. In a broad sense, most OC is characterized as immune cold, which may affect the efficacy of ICB treatment, but more explicit evidences are needed to systematically parse the tumor immune microenvironment (TIME) of OC[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. And a recent study shows that targeting St3gal3 to block aberrant sialylation in OC cells can alternate the TIME by converting the M2-like Macrophages to an M1-like phenotype and thus enhance the infiltration of functional CD8\u0026thinsp;+\u0026thinsp;T cells, improving the effectiveness of ICB treatment[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. This has shed light on the immunotherapy in OC, and highlights the contribution of the interaction between cancer cells and immune cells.\u003c/p\u003e \u003cp\u003eSmall GTPase plays an important role as molecular switches in cells, profoundly affecting the biological processes[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], and emerging studies point to its role in immunity. And there have been studies disclosing the correlation between small GTPase and the regulation of Macrophages[\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Moreover, in OC, perturbation of small GTPase contributes to a more metastatic, invasive and migrative ovarian cancer cells[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Simultaneously, treatments targeting small GTPase have been raised[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e] and great breakthrough has been made[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. These advances emphasize the necessity and urgency to explore the landscape of small GTPase in OC.\u003c/p\u003e \u003cp\u003eIn this study, we explored the transcriptional diversity between platinum-sensitive and resistant samples. And based on these differentially expressed genes, we divided the TCGA-OV cohort into two clusters, which had resembling gene expression features. Patients in the cluster with platinum-resistant features had significantly worse overall survival. We further detected the DEGs between the two clusters, and established a platinum-resistance related risk score using small GTPase binding related genes, which had terrific value in the prediction of OS in HGSOC cohorts. Combining with clinical information, a nomogram was constructed, and had robust efficacy. We further explored that SPP1\u0026thinsp;+\u0026thinsp;M2-like Macrophages had the most expression level of risky small GTPase binding related genes with single-cell RNA and spatial transcriptomics data. Moreover, NDRG1 and ABCA1 were the hub genes for SPP1\u0026thinsp;+\u0026thinsp;Macrophages in affecting OC patient survival.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData acquisition\u003c/h2\u003e \u003cp\u003eThe RNA sequencing data included in this study were collected from the TCGA-OV cohort (n\u0026thinsp;=\u0026thinsp;377), and three micro array data based on GPL570 platform were obtained from GSE9899 (n\u0026thinsp;=\u0026thinsp;167)[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e], GSE63885 (n\u0026thinsp;=\u0026thinsp;70)[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] and GSE26193 (n\u0026thinsp;=\u0026thinsp;79)[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e] with clinical information. Only patients with the diagnosis of serous ovarian cancer were taken into account. The RNA sequencing data of platinum-sensitive and resistant patients were collected from GSE160626[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], with 10 platinum-sensitive and 8 platinum-resistant samples from OC patients.\u003c/p\u003e \u003cp\u003eThe single-cell RNA sequencing data were obtained from recent research containing 7 platinum-sensitive and 3 platinum-resistant samples[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. While the spatial transcriptomics (ST) data were available in GSE211956, with 3 good response, 2 partial response and 3 poor response samples[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eDifferentially expressed gene (DEG) detection\u003c/h2\u003e \u003cp\u003eRegarding the transcripts per kilobase million (TPM) data in GSE160624, we utilized a Wilcoxon test to detect DEGs. The significant criteria were set as an absolute log fold change (FC) greater than 1 and a p-value smaller than 0.05 (see Supplementary materials Table\u0026nbsp;1).\u003c/p\u003e \u003cp\u003eFor raw counts data in the TCGA-OV cohort, three algorithms including \u003cem\u003eLimma\u003c/em\u003e[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], \u003cem\u003eedgeR\u003c/em\u003e[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] and \u003cem\u003eDESeq2\u003c/em\u003e[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e] were applied, the initial filtering criteria were set as an absolute logFC greater than 1 and a p-value smaller than 0.05, and the intersecting genes of three algorithms were considered as significant DEGs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eConsensus clustering\u003c/h2\u003e \u003cp\u003eThe consensus clustering was conducted with cluster algorithm set as \u003cem\u003ek-means\u003c/em\u003e, using 26 platinum-resistant related genes (genes with a corresponding Ensembl ID, not mitochondrial genes) based on the \u003cem\u003eConsensusClusterPlus\u003c/em\u003e[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] R packages. The optimal k value was determined by the minimal proportion of ambiguous clustering (PCA) value.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eGene enrichment analysis\u003c/h2\u003e \u003cp\u003eThe Gene Ontology (GO) analysis was conducted with \u003cem\u003eenrichGO\u003c/em\u003e command in R packages \u003cem\u003eclusterProfiler\u003c/em\u003e[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. And the gene set enrichment analysis (GSEA) was applied with \u003cem\u003eGSEA\u003c/em\u003e command with the HALLMARK and KEGG gene sets.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eSurvival analysis\u003c/h2\u003e \u003cp\u003eThe Kaplan-Meier (KM) analysis was conducted with R packages \u003cem\u003esurvival\u003c/em\u003e, and the optimal cut-off value was determined using R packages \u003cem\u003esurvminer\u003c/em\u003e. A p-value smaller than 0.05 was considered as statistically significant.\u003c/p\u003e \u003cp\u003eThe KM analysis results of ABCA1 and NDRG1 were obtained in KM-plotter website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://kmplot.com/analysis/\u003c/span\u003e\u003cspan address=\"https://kmplot.com/analysis/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e)[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eIntegrated machine learning method\u003c/h2\u003e \u003cp\u003eA total of 101 machine learning methods including Random Survival Forest (RSF), Enet, StepCox, CoxBoost, plsRCox, SuperPC, GBM, survival SVM, Least Absolute Shrinkage and Selection Operator (LASSO) and Ridge, and their combinations were applied in this study. Specifically, when combining two algorithms, the first algorithm was used to select variables accordingly and the second algorithm was utilized to build the predicting model with filtered variables.\u003c/p\u003e \u003cp\u003eFor RSF, the optimal mtry value was determined by method of exhaustion. And the relative importance of variables was obtained with subsample command in R package \u003cem\u003erandomForestSRC\u003c/em\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eEvaluation of immune signatures\u003c/h2\u003e \u003cp\u003eA total of 5 immune algorithms including CIBERSORT, MCPcounter, EPIC, TIMER and quantiseq were applied to evaluate the immune cell infiltration landscape in the TCGA-OV cohort using R package \u003cem\u003eIOBR\u003c/em\u003e[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. The immune signatures including stimulatory, inhibitory and antigen presenting molecules were summarized from previous research.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eProcessing of the single-cell RNA sequencing data\u003c/h2\u003e \u003cp\u003eThe processed data were obtained from recent research[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e], and to better integrate data from different samples, we performed the reciprocal PCA (RPCA) integration (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe signature score was calculated with \u003cem\u003eAddModuleScore\u003c/em\u003e command in R package \u003cem\u003eSeurat\u003c/em\u003e[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eFor the detection of DEGs, the criteria were set as a \u003cem\u003emin.pct\u003c/em\u003e greater than 0.2 and a logFC greater than 0.2.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eCell-cell communication analysis\u003c/h2\u003e \u003cp\u003eThe intra-cellular communication patterns were detected using R package \u003cem\u003eCellChat\u003c/em\u003e[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eProcessing of the ST data\u003c/h2\u003e \u003cp\u003eSpots in the ST data were firstly filtered with number of features greater than 300 and percentage of mitochondrial genes smaller than 0.3 as criteria. Then the normalized RNA value was obtained by the \u003cem\u003eSCTransform\u003c/em\u003e command. The signature score was calculated using \u003cem\u003eAddModuleScore\u003c/em\u003e command.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eAll data processing, statistical analysis, and plotting were performed in R 4.3.1 software. Correlations between two continuous variables were calculated via Sperman\u0026rsquo;s correlation analysis. The chi-squared test was used to compare categorical variables. The univariate and multivariate Cox regression analyses were conducted with R package \u003cem\u003esurvival\u003c/em\u003e. The receiver operating characteristic curve (ROC) was implemented using R package \u003cem\u003etimeROC\u003c/em\u003e, and the time-dependent ROC were calculated with R package \u003cem\u003ereportROC\u003c/em\u003e with a 1-month interval. All statistical tests performed were two-sided. A p-value smaller than 0.05 was considered as statistically significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eConsensus clustering with platinum-resistance related gene signatures defined two clusters in the TCGA-OV cohort\u003c/h2\u003e \u003cp\u003eFrom a previous RNA sequencing data on platinum-sensitive and resistant samples, we defined a total of 50 significant platinum-resistance related gene signatures (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA and Supplementary materials Table\u0026nbsp;1). After filtering mitochondrial genes and mapping to the TCGA-OV cohort, 26 genes were finally selected for the consensus clustering. Based on the PAC results, k\u0026thinsp;=\u0026thinsp;2 was considered optimal with least PAC (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). And the clustering results were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC. The summarized clinical status and specified expression patterns of two clusters were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD, where cluster 2 showed obvious elevation of platinum-resistance related genes. A total number of 4218 differentially expressed genes (DEGs) was detected between C2 and C1 (see Supplementary materials Table\u0026nbsp;2\u0026ndash;5). Using a GO enrichment analysis, the DEGs between two clusters were specifically enriched in small GTPase and GTPase activity and in the molecular function related to GTPase binding and small GTPase binding (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). Based on the KM results, there was significant survival advantages in the cluster 1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF), while there were no significant differences in the distribution of tumor grading and clinical staging between two clusters (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eG), as most patients suffered from high-graded and late-staged ovarian cancer. Moreover, proliferating related signatures TUBA4A and PROM2 were significantly higher in the cluster 2, and PROM2 was also proved to correlate with cancer stem cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eH).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eEstablishment of the platinum-resistance related risk score (PRRS) and nomogram with machine learning methods\u003c/h2\u003e \u003cp\u003eA total of 108 DEGs from two clusters within the GO small GTPase binding molecular function gene list were selected to build the PRRS. As was shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, in the TCGA-OV cohort, 101 kinds of prediction models were fitted and the C-index was calculated among 4 cohorts for validation. The RSF method was considered optimal with highest mean C-index. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB illustrated the top 20 important variables based on the RSF results. Based on the PRRS, in the combinate META-OV cohort and a single TCGA-OV cohort, with medium cut-off, patients with higher PRRS had significantly worse overall survival (OS) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). And with optimal cut-off, PRRS in all the validation cohorts significantly represented worse survival (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE-\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG). \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo test the efficacy of PRRS, we conducted univariate and multivariate Cox regression analyses. In the univariate Cox regression results, RiskScore, together with Grade 3/4 when compared with G1/2, Stage III/IV when compared with Stage I/II exhibited a Hazard ration (HR) greater than one, meaning they were risky factors for ovarian cancer (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). And in the multivariable Cox regression results, the RiskScore and Stage III/IV remained HR greater than one, indicating their robust predicting value for clinical outcomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). In order to build the bridge between the PRRS and clinical indicators, a nomogram was established (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). And the calibration curves of 1-year, 3-year and 5-year survival demonstrated that the nomogram had excellent predicting accuracy (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD). Moreover, the area under curve (AUC) value was 0.83 for 1 year, 0.848 for 3 years, and 0.838 for 5 years, indicating the terrific performance of the nomogram (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE). We also conducted the time dependent AUC changes at 1 month interval, and discovered that the nomogram and PRRS had great advantages as compared with Grade and Stage to predict the patient OS, and the nomogram was slightly more superb in predicting long time OS when compared with a single PRRS (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF). And the decision curve analysis (DCA) results also supported the above conclusions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG). These results jointed demonstrated that PRRS and a PRRS based nomogram had great clinical significances. \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eThe immune landscape and mutation landscape between two risk subgroups and their biological pathway differences\u003c/h2\u003e \u003cp\u003eTo comprehensively evaluate the immune landscape between two risk subgroups, 5 immune algorithms including CIBERSORT, MCPcounter, EPIC, TIMER and quantiseq were conducted based on the TCGA cohort. The summarized heatmap was shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA. And judging from a holistic point of view, myeloid lineages including Macrophages M2 in CIBERSORT and quantiseq, Monocytic lineage in MCPcounter, and DC in timer were significantly higher in the risk high subgroup, and neutrophils in MCPcounter and TIMER represented higher infiltration levels in the risk high subgroup. And the spearman correlation results proved evidence for the positive coupling of PRRS with above immune cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB-\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eG). Regarding stimulatory molecules, CD40LG was higher in the risk low subgroup, while ENTPD1 and ICOSLG had higher RNA level in the risk high subgroup. Inhibitory molecules, including ARG1, CD276, TGFB1 and VSIR had significant higher expression in the risk high subgroup. And there were no significant differences in antigen presenting molecules among two subgroups (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eH). \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThen we explored the biological pathway differences using DEGs from the Limma results. The gene set enrichment analysis (GSEA) results of HALLMARK gene set revealed that multiple tumor progression related hallmarks, including the MYC targets, PI3K AKT MTOR signaling, together with metabolic hallmarks, such as the Adipogenesis, glycolysis and fatty acid metabolism, were significantly elevated in the risk high subgroup, which indicated the more aggressive and proliferative phenotypes in the risk high subgroup (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). And the GO results also validated that the DEGs were enriched in metabolic pathways and pathways in cancer (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). Moreover, we explored the mutation landscape between two subgroups, while TP53 and TTN were the most common mutation in both groups, other mutations had different distributions. For example, the risk high subgroup had more CSMD3, MUC16 and USH2A mutations, while the risk low subgroup had more AHNAK, FAT3 and HUWE1 mutations (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC, \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD). \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eExploring the PRRS signatures with single-cell RNA sequencing data\u003c/h2\u003e \u003cp\u003eTo further determined how PRRS signatures affected the tumor microenvironment (TME), a total of 10 single-cell RNA data from patients with high-grade serous OC (HGSOC) was obtained from recent research[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e], and the number of filtered cells and patients\u0026rsquo; response to chemotherapy were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA (see Supplementary materials Table\u0026nbsp;6). A total of 9 major cell types were determined (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB) with canonical cell markers (Cancer cells: KRT18, CD24; Immune cells: PTPRC; T cells: CD3E; NK cells: GNLY; B/Plasma cells: CD79A; Myeloid cells: LYZ; Macrophages: C1QA; DC: C1DC; Endothelial cells: VWF; Fibroblast: COL1A2, DCN) (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). And the top 5 markers for each major cell type were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC. We calculated the ratio of observed to expected cell numbers (\u003cem\u003eRo/e\u003c/em\u003e) between resistant and sensitive patients, and consistent with the cell deconvolution results, the Macrophages were more enriched in the resistant samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD). We then calculated the PRRS signature scores, and discovered that the cells in the resistant samples compared with the sensitive samples, and especially high in the Macrophages in the resistant samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eE), meaning that PRRS signature genes most correlated with Macrophages. We further divided the Macrophages into 7 subgroups (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eF), and the comparing results indicated the subgroup Macro4 in the resistant subgroup had highest PRRS signature score among other subgroups (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eG). \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eCompared with Macro4 sensitive cells, SPP1 had remarkably higher expression in the Marco4 resistant cells, which was known markers for M2 Macrophages (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA). These elevated genes were enriched in inflammatory response and angiogenesis, providing evidence for their pro-tumor role (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB). Using the cell chat, we further explored the outgoing signaling patterns and incoming signaling patterns (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC, \u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eD), illustrating the special communication patterns in Macro4 resistant cells. Figure\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eE showed the significant ligand-receptor pairs, and Macro4 resistant cells had the strongest communication probability of SPP1 signaling pathways among other cells. And Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eF illustrated the specific ligands and receptors among cell types. \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eNDRG1 and ABCA1 were the hub genes for SPP1\u0026thinsp;+\u0026thinsp;Macrophages in affecting OC patient survival\u003c/h2\u003e \u003cp\u003eAmong these modeled genes of PRRS, NDRG1 and ABCA1 were also markers for Macro4 resistant cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA). A higher level of ABCA1 and NDRG1 (DRG1) represented worse survival (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB-\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eE). Based on ST results, we explored the distribution patterns and gene expression patterns of samples from HGSOC patients who received taxane- and platinum-based neoadjuvant chemotherapy. In the poor response group, there were specific area co-expressing ABCA1, NDRG1 and had especially higher Macro4 resistant signature score, specifically in P1 and P8 (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eF). While in the good response and partial response group, the co-expressing patterns was low and insufficient (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eG-\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eH). These results also proved evidence for the high RNA level of ABCA1 and NDRG1 in OC, paving way for their significant role in OC.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe acquired resistance to chemotherapy has always been a hot topic in the management of OC patients. With its high reoccurrence rate and severely high lethality, resistant OC after relapse normally results in a poor overall survival[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. The current assessment systems for serous OC failed to distinguish high risk patients, as most of them were diagnosed with advanced HGSOC. Though current research managed to build a scoring system for OC[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e] and explore the risky cell types[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e], and some attempted to build the bridge between drug resistance and patient survival[\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e], a more efficient model was of necessity. In our study, we identified platinum-resistant gene signatures, and the clustering results indicated a platinum-resistance resembling cluster, which had significantly worse OS. Interestingly, these DEGs in the platinum-resistance resembling cluster were enriched in small GTPase related gene sets. As a result, via integrated machine learning methods, we built a PRRS model based on these small GTPase related DEGs, which had ideal and stable performances among validation cohorts. Combined with current clinical indexes, a nomogram was produced and exhibited terrific predicting performances in predicting patient survival, which had a settled AUC value greater than 0.8, meaning high accuracy and sensitivity.\u003c/p\u003e \u003cp\u003eWith the advances in sequencing technology and the wide application of single-cell RNA sequencing and ST technology, emerging studies attempted to parse the OC in a more nuanced perspective and highlighted the significances of TIME[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Using deconvolution methods, we discovered a robust correlation with PRRS and M2 Macrophages. And the results were validated by the single-cell RNA sequencing data, where SPP1\u0026thinsp;+\u0026thinsp;M2-like Macrophages had the most expression of small GTPase related signatures, indicating a vital role of small GTPase molecules in regulating M2-like Macrophages. Via cell chat, we discovered that the specific SPP1\u0026thinsp;+\u0026thinsp;Macrophages had unique intracellular communication patterns, especially their dominant contributions in the SPP1 signaling pathway.\u003c/p\u003e \u003cp\u003eIn the last section of our study, we identified ABCA1 and NDRG1 (DRG1) as the hub genes for SPP1\u0026thinsp;+\u0026thinsp;Macrophages in affecting patient OS. Based on ST results, we discovered the co-localization of SPP1\u0026thinsp;+\u0026thinsp;Macrophages markers with ABCA1 and NDRG1. A previous research has also discovered ABCA1 mediates the acquired resistance in ovarian cancer cells, and correlates with poor OS[\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Acting as an important myeloid leukocyte activator, the role of NDRG1 in cancer was critical, and in ovarian cancer, knockdown of NDRG1 results in enhanced tumor cell adhesion, migration and invasion activities[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. But the role of ABCA1 and NDRG1 in regulating Macrophages in OC remained unclear. The present research showed the important role of ABCA1 and NDRG1 in SPP1\u0026thinsp;+\u0026thinsp;Macrophages, and deserved future research.\u003c/p\u003e \u003cp\u003eIn conclusion, our study identified a cluster of patients with platinum-resistant features, which had comparable shorter OS. We discovered that DEGs in the platinum-resistance resembling clusters were correlated with small GTPase. A PRRS model was built accordingly and was competent in detecting high risk patients among cohorts. Moreover, we firstly discussed the possible significance of ABCA1 and NDRG1 in SPP1\u0026thinsp;+\u0026thinsp;Macrophages, providing new perspectives for future research.\u003c/p\u003e \u003cp\u003eThis work had its limitations, the PRRS model and nomogram needed to be fully optimized with more clinical sample. And the exact mechanisms of ABCA1 and NDRG1 in regulating Macrophages required more in vitro and in vivo experiments.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cstrong\u003eHGSOC\u003c/strong\u003e: High grade serous ovarian cancer\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTIME\u003c/strong\u003e: Tumor immune microenvironment\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOS\u003c/strong\u003e: Overall survival\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eROC\u003c/strong\u003e: Receiver operating characteristic curve\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDCA\u003c/strong\u003e: Decision curve analysis\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePRRS\u003c/strong\u003e: Platinum-resistance related risk score\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eICB\u003c/strong\u003e: Immune checkpoint blockers\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eST\u003c/strong\u003e: Spatial transcriptomics\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDEG\u003c/strong\u003e: Differentially expressed gene\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTPM\u003c/strong\u003e: Transcripts per kilobase million\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFC\u003c/strong\u003e: Fold change\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGSEA\u003c/strong\u003e: Gene set enrichment analysis\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eKM\u003c/strong\u003e: Kaplan-Meier\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRSF\u003c/strong\u003e: Random Survival Forest\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLASSO\u003c/strong\u003e: Least Absolute Shrinkage and Selection Operator\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRPCA\u003c/strong\u003e: Reciprocal PCA\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePAC\u003c/strong\u003e: Proportion of ambiguous clustering\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGO\u003c/strong\u003e: Gene Ontology\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBP\u003c/strong\u003e: Biological process\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMF\u003c/strong\u003e: molecular function\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRo/e\u003c/strong\u003e: Ratio of observed to expected cell numbers\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data included in this study is available in the TCGA and GEO databases. All codes used for analyses are available on reasonable request\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNone.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHZ, WK and YZ proposed and designed this study. YZ, YW and WK conducted most of the formal analyses. YZ and SW made contributions to the embellishment of figures and wrote this manuscript. LB and RZ collected the data used for analyses. YZ and YW reviewed the data. All authors confirmed the final version for submission.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBowtell DD, B\u0026ouml;hm S, Ahmed AA, Aspuria PJ, Bast RC Jr., Beral V, Berek JS, Birrer MJ, Blagden S, Bookman MA, et al. Rethinking ovarian cancer II: reducing mortality from high-grade serous ovarian cancer. Nat Rev Cancer. 2015;15(11):668\u0026ndash;79.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatulonis UA, Sood AK, Fallowfield L, Howitt BE, Sehouli J, Karlan BY. Ovarian cancer. Nat Reviews Disease Primers. 2016;2(1):16061.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang L, Wang X, Zhu X, Zhong L, Jiang Q, Wang Y, Tang Q, Li Q, Zhang C, Wang H, et al. Drug resistance in ovarian cancer: from mechanism to clinical trial. Mol Cancer. 2024;23(1):66.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePawłowska A, Rekowska A, Kuryło W, Pańczyszyn A, Kotarski J, Wertel I. Current Understanding on Why Ovarian Cancer Is Resistant to Immune Checkpoint Inhibitors. Int J Mol Sci 2023, 24(13).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang ZB, Zhang X, Fang C, Liu XT, Liao QJ, Wu N, Wang J. Immunotherapy and the ovarian cancer microenvironment: Exploring potential strategies for enhanced treatment efficacy. Immunology; 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao K, Zhang G, Yang M, Wang Y, He M, Zhang C, Huang Y, Lu J, Liu H. Attenuation of Sialylation Augments Antitumor Immunity and Improves Response to Immunotherapy in Ovarian Cancer. Cancer Res. 2023;83(13):2171\u0026ndash;86.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJohnson DS, Chen YH. Ras family of small GTPases in immunity and inflammation. Curr Opin Pharmacol. 2012;12(4):458\u0026ndash;63.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong S, Cong W, Zhou S, Shi Y, Dai W, Zhang H, Wang X, He B, Zhang Q. Small GTPases: Structure, biological function and its interaction with nanoparticles. Asian J Pharm Sci. 2019;14(1):30\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCai Y, Sukhova GK, Wong HK, Xu A, Tergaonkar V, Vanhoutte PM, Tang EH. Rap1 induces cytokine production in pro-inflammatory macrophages through NFκB signaling and is highly expressed in human atherosclerotic lesions. Cell cycle (Georgetown Tex). 2015;14(22):3580\u0026ndash;92.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarvalho BC, Oliveira LC, Rocha CD, Fernandes HB, Oliveira IM, Le\u0026atilde;o FB, Valverde TM, Rego IMG, Ghosh S, Silva AM. Both knock-down and overexpression of Rap2a small GTPase in macrophages result in impairment of NF-κB activity and inflammatory gene expression. Mol Immunol. 2019;109:27\u0026ndash;37.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang S, Chen T, Yang M, Wang L, Yu Z, Xie B, Qian C, Xu S, Li N, Cao X, et al. Extracellular calcium elicits feedforward regulation of the Toll-like receptor-triggered innate immune response. Cell Mol Immunol. 2017;14(2):180\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao L, Zheng M, Guo Q, Nie X, Li X, Hao Y, Liu J, Zhu L, Lin B. Downregulation of Rab23 inhibits proliferation, invasion, and metastasis of human ovarian cancer. Int J Biochem Cell Biol. 2019;116:105617.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo J, Han X, Li J, Li Z, Yi J, Gao Y, Zhao X, Yue W. Single-cell transcriptomics in ovarian cancer identify a metastasis-associated cell cluster overexpressed RAB13. J translational Med. 2023;21(1):254.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrieto-Dominguez N, Parnell C, Teng Y. Drugging the Small GTPase Pathways in Cancer Treatment: Promises and Challenges. Cells 2019, 8(3).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYin G, Huang J, Petela J, Jiang H, Zhang Y, Gong S, Wu J, Liu B, Shi J, Gao Y. Targeting small GTPases: emerging grasps on previously untamable targets, pioneered by KRAS. Signal Transduct Target Therapy. 2023;8(1):212.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTothill RW, Tinker AV, George J, Brown R, Fox SB, Lade S, Johnson DS, Trivett MK, Etemadmoghadam D, Locandro B, et al. Novel molecular subtypes of serous and endometrioid ovarian cancer linked to clinical outcome. Clin cancer research: official J Am Association Cancer Res. 2008;14(16):5198\u0026ndash;208.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLisowska KM, Olbryt M, Dudaladava V, Pamuła-Piłat J, Kujawa K, Grzybowska E, Jarząb M, Student S, Rzepecka IK, Jarząb B, et al. Gene expression analysis in ovarian cancer - faults and hints from DNA microarray study. Front Oncol. 2014;4:6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMateescu B, Batista L, Cardon M, Gruosso T, de Feraudy Y, Mariani O, Nicolas A, Meyniel JP, Cottu P, Sastre-Garau X, et al. miR-141 and miR-200a act on ovarian tumorigenesis by controlling oxidative stress response. Nat Med. 2011;17(12):1627\u0026ndash;35.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu Y, Xu M, Feng Z, Wu H, Wu J, Ha X, Wu Y, Chen S, Xu F, Wen H, et al. AUF1-induced circular RNA hsa_circ_0010467 promotes platinum resistance of ovarian cancer through miR-637/LIF/STAT3 axis. Cell Mol Life Sci. 2023;80(9):256.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng X, Wang X, Cheng X, Liu Z, Yin Y, Li X, Huang Z, Wang Z, Guo W, Ginhoux F, et al. Single-cell analyses implicate ascites in remodeling the ecosystems of primary and metastatic tumors in ovarian cancer. Nat cancer. 2023;4(8):1138\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDenisenko E, de Kock L, Tan A, Beasley AB, Beilin M, Jones ME, Hou R, Muir\u0026iacute; D, Bilic S, Mohan G, et al. Spatial transcriptomics reveals discrete tumour microenvironments and autocrine loops within ovarian cancer subclones. Nat Commun. 2024;15(1):2860.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRitchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinf (Oxford England). 2010;26(1):139\u0026ndash;40.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLove MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinf (Oxford England). 2010;26(12):1572\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGyőrffy B. Transcriptome-level discovery of survival-associated biomarkers and therapy targets in non-small-cell lung cancer. Br J Pharmacol. 2024;181(3):362\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng D, Ye Z, Shen R, Yu G, Wu J, Xiong Y, Zhou R, Qiu W, Huang N, Sun L, et al. IOBR: Multi-Omics Immuno-Oncology Biological Research to Decode Tumor Microenvironment and Signatures. Front Immunol. 2021;12:687975.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHao Y, Hao S, Andersen-Nissen E, Mauck WM 3rd, Zheng S, Butler A, Lee MJ, Wilk AJ, Darby C, Zager M, et al. Integrated analysis of multimodal single-cell data. Cell. 2021;184(13):3573\u0026ndash;e35873529.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12(1):1088.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarchetti C, De Felice F, Romito A, Iacobelli V, Sassu CM, Corrado G, Ricci C, Scambia G, Fagotti A. Chemotherapy resistance in epithelial ovarian cancer: Mechanisms and emerging treatments. Sem Cancer Biol. 2021;77:144\u0026ndash;66.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang J, Wang C, Zhang Y, Cheng S, Wu M, Gu S, Xu S, Wu Y, Wang Y. A novel autophagy-related gene signature associated with prognosis and immune microenvironment in ovarian cancer. J ovarian Res. 2023;16(1):86.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang J, Zhang X, Zheng F, Yang Q, Bi F. Mitophagy-related long non-coding RNA signature predicts prognosis and drug response in Ovarian Cancer. J ovarian Res. 2023;16(1):177.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlbrecht S, Busschaert P, Qian J, Vanderstichele A, Loverix L, Van Gorp T, Van Nieuwenhuysen E, Han S, Van den Broeck A, Coosemans A, et al. High-grade serous tubo-ovarian cancer refined with single-cell RNA sequencing: specific cell subtypes influence survival and determine molecular subtype classification. Genome Med. 2021;13(1):111.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen L, Gao W, Lin L, Sha C, Li T, Chen Q, Wei H, Yang M, Xing J, Zhang M, et al. A methylation- and immune-related lncRNA signature to predict ovarian cancer outcome and uncover mechanisms of chemoresistance. J ovarian Res. 2023;16(1):186.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang W, Lokman NA, Noye TM, Macpherson AM, Oehler MK, Ricciardelli C. ABCA1 is associated with the development of acquired chemotherapy resistance and predicts poor ovarian cancer outcome. Cancer drug Resist (Alhambra Calif). 2021;4(2):485\u0026ndash;502.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao G, Chen J, Deng Y, Gao F, Zhu J, Feng Z, Lv X, Zhao Z. Identification of NDRG1-regulated genes associated with invasive potential in cervical and ovarian cancer cells. Biochem Biophys Res Commun. 2011;408(1):154\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang B, Li J, Ye Z, Li Z, Wu X. N-myc downstream regulated gene 1 acts as a tumor suppressor in ovarian cancer. Oncol Rep. 2014;31(5):2279\u0026ndash;85.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Ovarian cancer, Platinum resistance, Small GTPase, Predicting model","lastPublishedDoi":"10.21203/rs.3.rs-4336933/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4336933/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eHigh grade serous ovarian cancer (HGSOC) had high lethality due to its high relapse rate and acquired drug resistance. The tumor immune microenvironment (TIME) of HGSOC was heterogeneous, and mostly immune cold. We aimed to build the bridge between platinum-resistance related signatures and patient overall survival (OS).\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThe RNA sequencing data from GSE160626 was used for extraction of platinum-resistance related genes. The TCGA-OV cohort were fitted into 101 kinds of machine learning methods, and the validation cohort included GSE9899, GSE63885 and GSE26193. Numerous methods including the Cindex, receiver operating characteristic curve (ROC), univariate and multivariate Cox regression, and the decision curve analysis (DCA) were applied to detect the performances of platinum-resistance related risk score (PRRS) and a PRRS based nomogram. The single-cell RNA sequencing data and Spatial Transcriptomics data were used to determine the risky cell types correlated with our PRRS.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eBased on platinum-resistance related genes, we conducted consensus clustering and defined a platinum-resistance resembling cluster, which had significantly shorter OS. And with DEGs related to small GTPase between two clusters, we established a PRRS and a PRRS based nomogram, which had excellent performances in predicting OS of serous ovarian cancer patients. We further determined SPP1\u0026thinsp;+\u0026thinsp;M2-like Macrophages were risky factors correlated with the PRRS, and determined ABCA1 and NDRG1 as the hub genes related to patient OS.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eSmall GTPase was a dominant feature of platinum-resistance resembling clusters. PRRS had terrific predicting value and correlated with SPP1\u0026thinsp;+\u0026thinsp;M2-like Macrophages.\u003c/p\u003e","manuscriptTitle":"Identification of platinum-resistance related small GTPase binding signatures to predict the prognosis of ovarian cancer by machine learning and integrated bioinformatic analyses","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-07 18:50:58","doi":"10.21203/rs.3.rs-4336933/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d1ada502-c5d9-4871-b074-d590a0445e2f","owner":[],"postedDate":"May 7th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-10-15T05:39:06+00:00","versionOfRecord":[],"versionCreatedAt":"2024-05-07 18:50:58","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4336933","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4336933","identity":"rs-4336933","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}