Integrating Liquid-Liquid Phase Separation and Tumor Microenvironment Regulation: A lncRNA-Based Prognostic Model for Pancreatic Cancer

Integrating Liquid-Liquid Phase Separation and Tumor Microenvironment Regulation: A lncRNA-Based Prognostic Model for Pancreatic Cancer | 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 Integrating Liquid-Liquid Phase Separation and Tumor Microenvironment Regulation: A lncRNA-Based Prognostic Model for Pancreatic Cancer Yaqing Wei, Xiguang Sun, Changjun Ding, Yifei Wang, Zheran Lu, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7246386/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 Introduction: Liquid-Liquid Phase Separation (LLPS), tumor microenvironment (TME), and long non-coding RNA (lncRNA) all have varying degrees of influence on the expression regulation of tumors. However, research on the association of these three in pancreatic cancer (PC) still requires further exploration. This study seeks to establish the relationships among these three themes through bioinformatics and to identify biomarkers that can predict the prognosis of PC patients. Methods Data sets from The Cancer Genome Atlas (TCGA) and International Cancer Genome Consortium (ICGC) are obtained from the UCSC platform. lncRNAs associated with the LLPS and TME gene sets are screened, and model lncRNAs are identified through comprehensive analysis conducted with least absolute shrinkage and selection operator (LASSO) regression and cox proportional hazards (COX) regression. Additionally, the predictive efficacy of the model lncRNAs is validated through multiple databases and cohorts. Furthermore, the expression of the model lncRNAs is validated at a biological level. Results A comprehensive analysis establishes an optimal combination consisting of 5 lncRNAs. The Kaplan–Meier curves and receiver operating characteristic (ROC) curves for each cohort demonstrates the superiority of the model lncRNAs characteristics. Additionally, the COX regression analysis of clinical characteristics and the analysis of mutation data further indicates the stability of the model lncRNAs. Furthermore, the expression levels of model lncRNAs in cell lines are consistent with the analysis results. Conclusion The model lncRNAs identified in this study, which are correlated with LLPS and TME, demonstrate significant potential as independent biomarkers for predicting the prognosis of PC patients. pancreatic cancer LLPS TME and lncRNA Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Introduction PC has been increasing in both incidence and mortality rates each year. It has been reported that PC ranks 12th in incidence among all cancers and 6th in cumulative mortality rates. 1 It is project that by 2030, PC will become the second leading cause of cancer-related deaths in the United States. 2 Studies indicate that the lifetime risk of developing PC worldwide is up to 0.89% (95%CI: 0.88–0.89), and the median overall survival (OS) among all stages of the disease is only 4 months. 3 – 5 This severe treatment burden imposes a heavy burden on socioeconomic development, making it urgent to explore potential prognostic biomarkers to guide treatment for PC. Previous studies have shown that the assessment of clinical risk factors for PC prognosis is unsatisfactory. Known risk factors do not sufficiently explain the development of PC. 6 , 7 However, in recent years, lncRNA has garnered widespread attention in oncology research. 8 lncRNA influences tumor occurrence and progression through various mechanisms, including gene expression, post-transcriptional regulation, and cell proliferation. 9 , 10 Particularly in PC, abnormal expression of lncRNA is closely associated with tumor aggressiveness, metastasis, and prognosis. 11 , 12 Additionally, the characteristics of the PC microenvironment are distinct, including hypoxia and high fibrosis. 13 , 14 lncRNA participates in the construction of this network by regulating immune cells and cancer-associated fibroblasts within the TME. 15 , 16 Notably, lncRNA can also influence the expression of the TME by regulating the extracellular matrix, metabolic status, and intercellular signaling. 17 It is worth noting that the close relationship between lncRNA and LLPS during tumor progression is one of the important factors affecting the TME. 18 , 19 lncRNA can form condensates in the nucleus through LLPS, and these condensates not only participate in the regulation of gene transcription but may also act as signal integrators to respond to changes in the microenvironment. 20 , 21 Normal biomolecular condensates ensure fundamental cellular functions, while their anomalous forms can lead to cellular dysfunction and potential tumor. 22 Studies have demonstrated that LLPS plays a crucial role in regulating tumor proliferation and metastasis. 23 , 24 A significant number of lncRNAs are localized on chromatin and often form RNA-clouds within specific nuclear regions to regulate gene expression. As major components of various membraneless structures, such as nucleolus and paraspeckles, lncRNAs are frequently imbalanced in cancer. 25 Recent research has shown that certain lncRNA subdomains can selectively combine to proteins like NONO/SFPQ, which dynamically oligomerize and recruit additional proteins through LLPS. 26 This process not only participates in gene expression regulation, RNA processing, and cellular stress responses but is also highly sensitive to cellular states and environmental signals. 27 , 28 In summary, this study will anchor biomarkers at the mechanistic level of tumors, screening lncRNA based on the LLPS and TME gene sets to give a reference for the prognosis of PC. Methods Data acquisition and integration The data included in this study originates from the TCGA and ICGC datasets on the UCSC platform. 165 tumor samples and 171 normal tissue samples originate from the GTEx-TCGA dataset. This dataset is employed for model construction and serves as an internal validation dataset. To enhance the stability of the model, Researchers also include 90 pancreatic tumor samples sourced from the ICGC dataset, which serve as an external validation dataset. Additionally, differential expression analysis of the GTEx-TCGA gene matrix is conducted with the R language software version 4.2.3 (log 2 FC > 1.0, FDR < 0.05, P- value < 0.05). Overall, a total of 5,901 genes and 200 lncRNAs are identified as significantly differentially expressed. Confirm the theme-related target gene and potential lncRNA Based on the GeneCards search engine, gene sets are retrieved with the keywords 'TME' and 'LLPS'. Following this, the intersection of these two gene sets is constructed. This intersected gene set is then combined with the 5,901 differentially expressed genes to identify those that are associated with both themes. Subsequently, the limma R package is applied to perform a correlation analysis between the identified lncRNAs and the intersected genes. As a result, 138 lncRNAs that are linked to both 'TME' and 'LLPS' are confirmed as potential lncRNAs for further analysis. Model construction and cohort grouping To test the effectiveness of the model lncRNA, 165 tumor samples from the GTEx-TCGA dataset are randomly divided into two independent cohorts: a training cohort and a validation cohort. The optimal model is constructed with COX regression, and LASSO regression. To assess the performance of the model lncRNA and reduce the risk of overfitting, 10-fold cross-validation is conducted along with 1,000 random perturbations applied to the training cohort. The formula is then applied to calculate the risk score, where lncRNA k = coef(lncRNA k )× expr(lncRNA k ), with "coef" representing the survival coefficient and "expr" indicating the gene expression level. Based on the median risk score, both the training and validation cohort, as well as the external validation dataset, are all classified into low-risk and high-risk subgroups for further analysis. $$\:\text{r}\text{i}\text{s}\text{k}\:\text{s}\text{c}\text{o}\text{r}\text{e}={\sum\:}_{\text{k}=1}^{\text{n}}\left({\text{l}\text{n}\text{c}\text{R}\text{N}\text{A}}^{\text{k}}\right)\:$$ Assess the survival prediction efficacy of model lncRNA The researchers conduct internal validation by comparing survival status and survival curves of risk subgroups within the training and validation cohorts. To further evaluate the predictive efficacy of model lncRNA in the internal dataset, time-dependent ROC curves and clinical characteristics ROC curves are employed. Building on this evaluation, survival analyses and time-dependent ROC curves for the risk subgroups are performed on the entire TCGA dataset. These analyses are also conducted in an external dataset to illustrate the efficacy of the model lncRNA. Furthermore, researchers compare survival differences in clinical characteristics, such as age, gender, tumor grade, and tumor stage, between low-risk and high-risk subgroups. Finally, the reliability of risk scores is confirmed through the construction of univariate and multivariate COX regression analysis, which develops the predictive capability of the model lncRNA. Construction of mutation data analysis To further illustrate the efficacy of the model lncRNA, this study extracts mutation data from samples in the TCGA database. Researchers perform correlation analyses to reveal differences in mutation data between the low-risk and high-risk subgroups. Following this, the study categorizes the samples into two groups based on the median tumor mutation burden (TMB) coefficient: low-TMB (L-TMB) and high-TMB (H-TMB). Subsequently, researchers construct survival analysis curves for the mutation subgroups. Finally, researchers correlate the mutation subgroups with the risk subgroups to conduct survival analysis. This analysis strengthens the evidence for the model lncRNAs effectiveness and stability. Model lncRNAs expression validation in cell lines In this study, four cell lines—hTERT-HPNE (HPDE), PANC-1, ASPC-1, and MIA-PACA-2—are selected for qPCR experiments. The primer sequences applied in the research can be found in supplementary materials. Each group of experiments is conducted in triplicate, and the results are averaged, with the standard deviation calculated to assess variability. The qPCR results are then statistically analyzed employing the 2 −ΔΔCT method and one-way chi-square analysis. Finally, the results are visualized with Prism software version 9.5.1. Results Selection of target gene and model lncRNA The research flowchart for this study is shown in Fig. 1 . Through the analysis of sample data from the GTEx-TCGA dataset, researchers discover 3469 genes that are differentially expressed (Supplementary Material-1) . Researchers then obtained 728 theme genes from the GeneCards website (Supplementary Material-2) . After intersecting the differentially expressed genes with the theme genes, researchers discover 72 target genes ( Fig. 2 A ) . From the correlation analysis results, researchers discover 138 lncRNAs linked to target genes, prompting the development of a heatmap showing the top 30 lncRNAs ( Fig. 2 B and Supplementary Material-3) . The training and validation cohorts are validated as independent datasets based on the chi-square method ( Table 1 ) . Analyzing the sample data from the training cohort, researchers initially identify 7 lncRNAs through univariate COX regression analysis ( Fig. 2 C ) . To validate the result, researchers subsequently perform LASSO analysis, conducting 1,000 random cycles ( Figs. 2 D and 2 E ) . Finally, through multivariate COX regression analysis, 5 lncRNAs are selected as model lncRNAs ( Fig. 3 A ) . Table 1 Basic information of the random cohort. Clinical Indicators Classification Total Validation Training P -value Age (Year) 65 77 39(48.75%) 38(46.91%) Gender Female 74 36(45.00%) 38(46.91%) 0.9319 Male 87 44(55.00%) 43(53.09%) Grade 1 25 11(13.75%) 14(17.28%) 0.4424 2 89 48(60.00%) 41(50.62%) 3 46 20(25.00%) 26(32.10%) 4 1 1(1.25%) 0(0%) Stage Ⅰ 17 10(12.50%) 7(8.64%) 0.6011 Ⅱ 137 68(85.00%) 69(85.19%) Ⅲ 3 1(1.25%) 2(2.47%) Ⅳ 4 1(1.25%) 3(3.70%) T staging T 1 6 5(6.25%) 1(1.23%) 0.3240 T 2 19 8(10.00%) 11(13.58%) T 3 133 66(82.50%) 67(82.72%) T 4 3 1(1.25%) 2(2.47%) N staging N 0 46 24(30.00%) 22(27.16%) 0.8225 N 1 115 56(70.00%) 59(72.84%) Establishment and expression differential analysis of model lncRNA The correlation analysis based on the model lncRNAs and differently expressed genes show that the model lncRNAs are related to many co-related genes in LLPS and TMU ( Fig. 3 B ) . Additionally, researchers assess the expression differences of model lncRNAs between tumor and normal samples, as well as within the subgroups of high-risk and low-risk ( Figs. 3 C and 3 D ) . The results indicate significant differences in the expression of model lncRNAs among samples. Furthermore, the results of model RNA are consistent with the multivariate COX result, with AL157392.3 and BX284668.2 as low-risk genes, while the other model lncRNAs are high-risk genes. Internal dataset survival analysis The analysis of risk subgroups reveals significant differences in survival outcomes in both the training and validation cohorts (Supplementary Materials-4,5) . Specifically, the prognosis for the low-risk subgroup is found to be significantly longer than that for the high-risk subgroup ( Figs. 4 A and 4 B ) . In both cohorts, the assessment of risk status indicates that the number of patients deaths clearly increases with the risk scores ( Figs. 4 C and 4 D ) . The area under the time-dependent ROC curve (AUC) for both datasets surpass 0.66, effectively demonstrating the prediction capability of model lncRNAs ( Figs. 4 E and 4 F ) . Furthermore, ROC curves for clinical characteristics indicate that the risk scores is superior to other clinical characteristics as an indicator of prognosis (AUC > 0.7) ( Figs. 4 G and 4 H ) . To further evaluate the predictive performance of the model, this study assesses the survival time in risk subgroups across various clinical categories (age, sex, grade, and stage, etc). The results highlight significant differences between the low-risk and high-risk subgroups in most clinical categories, thereby further confirming the predictive accuracy of the model lncRNAs ( Figs. 5 A-L ) . External database validation This study further analyzes the differences in risk subgroups among samples from the TCGA and ICGC databases (Supplementary Materials-6,7) . Survival curves indicate that the high-risk subgroup has a shorter survival time compared to the low-risk subgroup ( Figs. 6 A and 6 B ) . Additionally, the assessment of survival status clearly demonstrates that there are fewer patients surviving in the high-risk subgroup ( Figs. 6 C and 6 D ) . The AUCs based on samples from both databases are basically greater than 0.6 ( Figs. 6 E and 6 F ) . While the time-dependent ROC curve within the ICGC dataset shows the lowest value of 0.597 during the first year, it is noteworthy that the predictive accuracy of the model exhibits an overall upward trend over time. Moreover, the AUC for clinical characteristics based on samples from both datasets exceeds 0.68, further confirming the predictive efficacy of the model lncRNAs ( Figs. 6 G and 6 H ) . Through verification analysis across different databases, this study evaluates the survival prediction capability of the model lncRNAs from multiple perspectives. Assessment of clinical features of model lncRNAs To further validate the predictive efficacy of the model lncRNAs for prognosis, this study conducts univariate and multivariate COX regression analyses employing clinical characteristics from samples in the TCGA and ICGC datasets. Both univariate and multivariate COX analyses provide consistent results, indicating that the risk score is significantly associated with patient prognosis ( Figs. 7 A-D ) . This demonstrates that the risk score can serve as an independent risk factor for clinical prognosis assessment, thereby further highlighting the stability of the model lncRNAs in evaluating prognostic efficacy. Additionally, the analysis indicates that tumor stage in ICGC dataset and tumor N staging in TCGA dataset also serve as independent factors. Mutation differences in risk subgroups To further compare the differences between the risk subgroups, this study analyzes mutation data of the samples (Supplementary Material-8) . Researchers obtain TMB values for the samples from the TCGA database. The analysis indicates a statistically significant difference in TMB values between the two risk subgroups, demonstrating that mutations vary between them (Fig. 8 A). Additionally, the correlation analysis result demonstrates a significant positive correlation between risk scores and TMB values ( Fig. 8 B ) . Specifically, this analysis shows that as the risk score increases, the TMB value also rises significantly. Furthermore, survival analysis results reveal clear differences in survival time in the subgroups ( Figs. 8 C and 8 D ) . This finding further highlights the sensitivity and stability of risk genes in predicting the prognosis status of the samples. qPCR validation of model lncRNAs expression The study analyzes the expression levels of model lncRNAs among different cell lines (Supplementary Materials-9, 10) . Specifically, the gene expression levels of the HPDE normal pancreatic ductal tissue cell line are selected as a reference. The protective genes AL157392.3 and BX284668.2 demonstrate significantly higher expression levels in the HPDE cell line compared to three other tumor cell lines ( Figs. 9 A and 9 B ) . This result is consistent with the results of the previous analysis. Furthermore, the high risk genes ( FAM83A-AS1 , LINC02257 , and TMEM105 ) also exhibit higher expression levels in the three tumor cell lines ( Figs. 9 C-E). In summary, the experimental results from qPCR are consistent with database analyses, which further confirms the accuracy and stability of the model lncRNAs at the biological level. Discussion Previous studies have shown that LLPS has a significant impact on the formation and regulation of lncRNA, and there is a mutual relationship between TME and LLPS. 29 , 30 Additionally, the regulatory role of lncRNAs in various cancers has been well confirmed. 31 , 32 Building on these insights, this study selects LLPS and TME as focal points, with lncRNAs as the research object, to explore its potential in predicting the prognosis of PC. Given that this study aims to explore the impact of lncRNAs expression related to TME and LLPS genes on the prognosis of PC patients. To ensure the validity of our results, samples with OS of fewer than 30 days are excluded from the analysis (may result from complications such as infections or surgical issues). This study employs a combined analysis of COX regression and LASSO regression to select an ideal lncRNA combination. Based on the analysis results, the optimal combination is established, which includes six lncRNAs: AL157392.3 , BX284668.2 , FAM83A-AS1 , LINC02257 , and TMEM105 . Notably, AL157392.3 and BX284668.2 in this combination act as protective genes, and their high expression is beneficial for patient prognosis. Furthermore, the validation of results from risk subgroup survival analyses across multiple databases confirms the superiority of this model. However, researchers observe no statistically significant differences in the Kaplan-Meier survival curves for clinical subgroups in the stage III−IV group. This lack of significance may arise from the limited sample size in this subgroup. 33 Additionally, the time-dependent and clinical characteristics ROC curves for risk subgroups further confirm the superiority of the model lncRNAs. Researchers also note that the superiority of the risk scores is diluted during the evaluation of data from different databases. It is thought by researchers that this finding may be linked to the small sample size in the ICGC dataset. Moreover, it is worth noting that analysis results indicate that the model lncRNAs can serve as independent risk factors for patient prognosis. If sufficient samples become available in the future, the research team will further validate the results. Additionally, the analysis related to TMB values indicates that the differences in survival times among samples within the independent mutation subgroup are not statistically significant. This implies that variations in TMB values have no significant impact on the survival time of these samples. However, the combined analysis of mutation subgroups with risk subgroups reveals significant statistical differences, further highlighting the validity and accuracy of the model lncRNAs. Furthermore, the expression levels of the model lncRNAs across the four cell line groups are consistent with analysis results, providing biological evidence that supports the reliability of the results. The lncRNAs AL157392.3 , FAM83A-AS1 , LINC02257 , and TMEM105 have been previously shown to have a positive correlation with the progression of various cancers, which aligns with the results of this study. 34 – 37 Limitation It is worth noting that, although the study validates the superiority of model lncRNAs from various angles, there are still several limitations inherent in the research. Undeniably, this study is retrospective in nature, which means that its results may be influenced by the intrinsic design of this research method. 38 – 40 To enhance the evaluation of model lncRNAs performance, researchers attempt to include more samples; however, corresponding lncRNAs expression information have not yet been obtained from gene matrices in existing databases. This is also the next direction for the research team, which will involve collecting more clinical samples to further explore and validate the risk signature. In summary, this study successfully identifies model lncRNAs associated with both LLPS and TME at the database level and biological level to construct a risk signature for predicting prognosis of PC patients. Conclusion This study associates the LLPS and TME gene sets through multi-database analysis and biological experiments, establishing a potential independent risk factor composed of lncRNAs to predict the prognosis of PC patients. Declarations Ethical approval The sample data involved in this study are all sourced from public database. The samples obtained from the public databases comply with the Helsinki Declaration and adhere to ethical standards. All uploaded data are publicly accessible and have received consent from the participants. Consent to participate Not applicable Consent for publication The manuscript is original and has not been published or accepted for publication elsewhere, either in whole or in part. Availability of data and material Pancreatic sample data are obtained from the UCSC platform (https://xenabrowser.net/datapages/); Gene sets are retrieved from the GenCard website (https://www.genecards.org/), and primer sequences are designed from the NCBI database (https://www.ncbi.nlm.nih.gov/). Competing interests The authors declare that this research is carried out without any commercial or financial relationships that could be interpreted as a potential conflict of interest. Funding This study was supported by the 2023 Youth Incubation Fund of The Second Hospital of Tianjin Medical University (Number: 2023ydey07). Author ’s Contributions YW, XS: writing; data analysis; and data visualization. CD, YW, ZL: data analysis. CZ, HY: review and editing. HH: review; editing; supervision. All authors have contributed to the publication of the manuscript and have agreed to its submission. Acknowledgement s Gratitude is extended to the UCSC platform for providing the public data cohorts from TCGA and ICGC; to GenCards and NCBI for their valuable search engines; and to all authors for their careful review and constructive feedback on this manuscript. References Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: a Cancer Journal For Clinicians . 2024;74(3):229-263. doi:10.3322/caac.21834 Rahib L, Wehner MR, Matrisian LM, Nead KT. Estimated Projection of US Cancer Incidence and Death to 2040. JAMA Netw Open . 2021;4(4):e214708. doi:10.1001/jamanetworkopen.2021.4708 Wang S, Zheng R, Li J, et al. Global, regional, and national lifetime risks of developing and dying from gastrointestinal cancers in 185 countries: a population-based systematic analysis of GLOBOCAN. Lancet Gastroenterol Hepatol . 2024;9(3):229-237. doi:10.1016/S2468-1253(23)00366-7 Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA: a Cancer Journal For Clinicians . 2024;74(1):12-49. doi:10.3322/caac.21820 Mackay TM, Latenstein AEJ, Augustinus S, et al. Implementation of Best Practices in Pancreatic Cancer Care in the Netherlands: A Stepped-Wedge Randomized Clinical Trial. JAMA Surg . 2024;159(4):429-437. doi:10.1001/jamasurg.2023.7872 Klein AP. Pancreatic cancer epidemiology: understanding the role of lifestyle and inherited risk factors. Nat Rev Gastroenterol Hepatol . 2021;18(7):493-502. doi:10.1038/s41575-021-00457-x Huang J, Lok V, Ngai CH, et al. Worldwide Burden of, Risk Factors for, and Trends in Pancreatic Cancer. Gastroenterology . 2021;160(3):744-754. doi:10.1053/j.gastro.2020.10.007 Nojima T, Proudfoot NJ. Mechanisms of lncRNA biogenesis as revealed by nascent transcriptomics. Nat Rev Mol Cell Biol . 2022;23(6):389-406. doi:10.1038/s41580-021-00447-6 Ahmad M, Weiswald L-B, Poulain L, Denoyelle C, Meryet-Figuiere M. Involvement of lncRNAs in cancer cells migration, invasion and metastasis: cytoskeleton and ECM crosstalk. J Exp Clin Cancer Res . 2023;42(1):173. doi:10.1186/s13046-023-02741-x Tan Y-T, Lin J-F, Li T, Li J-J, Xu R-H, Ju H-Q. LncRNA-mediated posttranslational modifications and reprogramming of energy metabolism in cancer. Cancer Commun (Lond) . 2021;41(2):109-120. doi:10.1002/cac2.12108 Hashemi M, Moosavi MS, Abed HM, et al. Long non-coding RNA (lncRNA) H19 in human cancer: From proliferation and metastasis to therapy. Pharmacol Res . 2022;184:106418. doi:10.1016/j.phrs.2022.106418 McCabe EM, Rasmussen TP. lncRNA involvement in cancer stem cell function and epithelial-mesenchymal transitions. Semin Cancer Biol . 2021;75:38-48. doi:10.1016/j.semcancer.2020.12.012 Tao J, Yang G, Zhou W, et al. Targeting hypoxic tumor microenvironment in pancreatic cancer. J Hematol Oncol . 2021;14(1):14. doi:10.1186/s13045-020-01030-w Pereira BA, Vennin C, Papanicolaou M, et al. CAF Subpopulations: A New Reservoir of Stromal Targets in Pancreatic Cancer. Trends Cancer . 2019;5(11):724-741. doi:10.1016/j.trecan.2019.09.010 Sempere LF, Powell K, Rana J, Brock AA, Schmittgen TD. Role of non-coding RNAs in tumor progression and metastasis in pancreatic cancer. Cancer Metastasis Rev . 2021;40(3):761-776. doi:10.1007/s10555-021-09995-x Gong Y, Gong D, Liu S, et al. Deciphering the role of NcRNAs in Pancreatic Cancer immune evasion and drug resistance: a new perspective for targeted therapy. Frontiers In Immunology . 2024;15:1480572. doi:10.3389/fimmu.2024.1480572 Shi M, Zhang R, Lyu H, et al. Long non-coding RNAs: Emerging regulators of invasion and metastasis in pancreatic cancer. J Adv Res . 2025;doi:10.1016/j.jare.2025.02.001 Tong X, Tang R, Xu J, et al. Liquid-liquid phase separation in tumor biology. Signal Transduct Target Ther . 2022;7(1):221. doi:10.1038/s41392-022-01076-x Guo Q, Shi X, Wang X. RNA and liquid-liquid phase separation. Noncoding RNA Res . 2021;6(2):92-99. doi:10.1016/j.ncrna.2021.04.003 Chen W, Li Y, Zhou Q, et al. The cancer-associated fibroblast facilitates YAP liquid-liquid phase separation to promote cancer cell stemness in HCC. Cell Commun Signal . 2025;23(1):238. doi:10.1186/s12964-025-02256-2 Somasundaram K, Gupta B, Jain N, Jana S. LncRNAs divide and rule: The master regulators of phase separation. Front Genet . 2022;13:930792. doi:10.3389/fgene.2022.930792 Roden C, Gladfelter AS. RNA contributions to the form and function of biomolecular condensates. Nat Rev Mol Cell Biol . 2021;22(3):183-195. doi:10.1038/s41580-020-0264-6 Mehta S, Zhang J. Liquid-liquid phase separation drives cellular function and dysfunction in cancer. Nat Rev Cancer . 2022;22(4):239-252. doi:10.1038/s41568-022-00444-7 Zheng L-W, Liu C-C, Yu K-D. Phase separations in oncogenesis, tumor progressions and metastasis: a glance from hallmarks of cancer. J Hematol Oncol . 2023;16(1):123. doi:10.1186/s13045-023-01522-5 Wang Y, Chen L-L. Organization and function of paraspeckles. Essays Biochem . 2020;64(6):875-882. doi:10.1042/EBC20200010 Zhao J, Xie W, Yang Z, et al. Identification and characterization of a special type of subnuclear structure: AGGF1-coated paraspeckles. FASEB J . 2022;36(6):e22366. doi:10.1096/fj.202101690RR Ren J, Zhang Z, Zong Z, Zhang L, Zhou F. Emerging Implications of Phase Separation in Cancer. Adv Sci (Weinh) . 2022;9(31):e2202855. doi:10.1002/advs.202202855 Yan X, Zhang M, Wang D. Interplay between posttranslational modifications and liquid‒liquid phase separation in tumors. Cancer Lett . 2024;584:216614. doi:10.1016/j.canlet.2024.216614 Nozawa R-S, Yamamoto T, Takahashi M, et al. Nuclear microenvironment in cancer: Control through liquid-liquid phase separation. Cancer Sci . 2020;111(9):3155-3163. doi:10.1111/cas.14551 Liu Y-T, Cao L-Y, Sun Z-J. The emerging roles of liquid-liquid phase separation in tumor immunity. Int Immunopharmacol . 2024;143(Pt 1):113212. doi:10.1016/j.intimp.2024.113212 Bin W, Yuan C, Qie Y, Dang S. Long non-coding RNAs and pancreatic cancer: A multifaceted view. Biomed Pharmacother . 2023;167:115601. doi:10.1016/j.biopha.2023.115601 Ashrafizadeh M, Rabiee N, Kumar AP, Sethi G, Zarrabi A, Wang Y. Long noncoding RNAs (lncRNAs) in pancreatic cancer progression. Drug Discov Today . 2022;27(8):2181-2198. doi:10.1016/j.drudis.2022.05.012 Zhou S, Shen C. Avoiding Definitive Conclusions in Meta-analysis of Heterogeneous Studies With Small Sample Sizes. JAMA Otolaryngol Head Neck Surg . 2022;148(11):1003-1004. doi:10.1001/jamaoto.2022.2847 Ho K-H, Huang T-W, Shih C-M, et al. Glycolysis-associated lncRNAs identify a subgroup of cancer patients with poor prognoses and a high-infiltration immune microenvironment. BMC Med . 2021;19(1):59. doi:10.1186/s12916-021-01925-6 Wang H, Ding Y, Zhu Q, et al. LncRNA FAM83A-AS1 promotes epithelial-mesenchymal transition of pancreatic cancer cells via Hippo pathway. Cell Cycle . 2023;22(12):1514-1527. doi:10.1080/15384101.2023.2216507 Park M-S, Jeong SD, Shin CH, et al. LINC02257 regulates malignant phenotypes of colorectal cancer via interacting with miR-1273g-3p and YB1. Cell Death Dis . 2024;15(12):895. doi:10.1038/s41419-024-07259-4 Yin Y, Sun Y, Yao H, et al. TMEM105 modulates disulfidptosis and tumor growth in pancreatic cancer via the β-catenin-c-MYC-GLUT1 axis. Int J Biol Sci . 2025;21(5):1932-1948. doi:10.7150/ijbs.104598 Ramgopal S, Benedetti J, Cotter JM. Performing a Multicenter Retrospective Study. Hosp Pediatr . 2025;15(2):e77-e82. doi:10.1542/hpeds.2024-008020 Greenland S, Senn SJ, Rothman KJ, et al. Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations. Eur J Epidemiol . 2016;31(4):337-350. doi:10.1007/s10654-016-0149-3 Kaji AH, Schriger D, Green S. Looking through the retrospectoscope: reducing bias in emergency medicine chart review studies. Ann Emerg Med . 2014;64(3):292-298. doi:10.1016/j.annemergmed.2014.03.025 Additional Declarations No competing interests reported. Supplementary Files 1.SupplementaryMaterials1.xlsx 2.SupplementaryMaterials2.xlsx 3.SupplementaryMaterials3.xlsx 4.SupplementaryMaterials4.xlsx 5.SupplementaryMaterials5.xlsx 6.SupplementaryMaterials6.xlsx 7.SupplementaryMaterials7.xlsx 8.SupplementaryMaterials8.xlsx 9.SupplementaryMaterials9.xlsx 10.SupplementaryMaterials10.xlsx 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-7246386","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":507851631,"identity":"64dc5c2c-7f69-4194-b1c3-c482ce8f9819","order_by":0,"name":"Yaqing Wei","email":"","orcid":"","institution":"The Second Hospital of Tianjin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yaqing","middleName":"","lastName":"Wei","suffix":""},{"id":507851632,"identity":"49760fd3-6c5a-4308-9dea-2207c277f4a2","order_by":1,"name":"Xiguang Sun","email":"","orcid":"","institution":"The Second Hospital of Tianjin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Xiguang","middleName":"","lastName":"Sun","suffix":""},{"id":507851633,"identity":"66f08e5b-7bdd-4376-9bf3-ed81005e3dd2","order_by":2,"name":"Changjun Ding","email":"","orcid":"","institution":"The Second Hospital of Tianjin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Changjun","middleName":"","lastName":"Ding","suffix":""},{"id":507851634,"identity":"bb6baaee-b4a4-4208-bd57-de121e2e72eb","order_by":3,"name":"Yifei Wang","email":"","orcid":"","institution":"The Second Hospital of Tianjin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yifei","middleName":"","lastName":"Wang","suffix":""},{"id":507851635,"identity":"290ef067-923a-4444-aded-aad518510a19","order_by":4,"name":"Zheran Lu","email":"","orcid":"","institution":"RWTH Aachen University","correspondingAuthor":false,"prefix":"","firstName":"Zheran","middleName":"","lastName":"Lu","suffix":""},{"id":507851636,"identity":"b9b39eab-075c-46d3-a51c-fe74a60adb81","order_by":5,"name":"Chenhui Zhang","email":"","orcid":"","institution":"The Second Hospital of Tianjin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Chenhui","middleName":"","lastName":"Zhang","suffix":""},{"id":507851637,"identity":"97d6d1ea-c8e3-450d-9b1b-5f0465ab9686","order_by":6,"name":"Hao Yao","email":"","orcid":"","institution":"The Second Hospital of Tianjin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Yao","suffix":""},{"id":507851638,"identity":"42c15c47-e143-4f4d-a38c-0b35a62508f3","order_by":7,"name":"Hao Huang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIiWNgGAWjYBACAxDB2CAhx8bffAAidIA4LRbG/BLHEhtI0VKROLMhx5A4LeYSyc8eft0hwbjhwJnvj262Mcjx3Uhg/FyAR4vljDRzY9kzEswGh3s3Nue2MRhL3khglp6Bz2E3EsykJdsk2AwOnAVrSdxwI4GNmQevlvRvIC08BgdyHoK01BOhJcdM8mObhIRkQw4jSEuCAUEtZ96USTO2SRgAA9lwds45CcOZZx42S+PVcjx9m+TPtrr6Nv7mB59zymzk+Y4nH/yMTwsIIDtDggEUTQQ0AJX8IKhkFIyCUTAKRjQAAIwsUPa/0FeZAAAAAElFTkSuQmCC","orcid":"","institution":"The Second Hospital of Tianjin Medical University","correspondingAuthor":true,"prefix":"","firstName":"Hao","middleName":"","lastName":"Huang","suffix":""}],"badges":[],"createdAt":"2025-07-29 19:38:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7246386/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7246386/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":90381580,"identity":"26d201f7-6b12-4195-add2-6c9de5bf56a0","added_by":"auto","created_at":"2025-09-02 06:48:44","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":4329665,"visible":true,"origin":"","legend":"\u003cp\u003eResearch flowchart. The arrows indicate the steps of research progress.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/2251098a16274f7407fbe699.png"},{"id":90383360,"identity":"17d58594-396d-4efd-9654-8a9220b28014","added_by":"auto","created_at":"2025-09-02 06:56:44","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1435998,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of target genes and lncRNAs. \u003cstrong\u003eA\u003c/strong\u003e: Volcano plot of 72 target genes. \u003cstrong\u003eB\u003c/strong\u003e: Heatmap of the top 30 target genes with the highest correlation from the correlation analysis results. \u003cstrong\u003eC\u003c/strong\u003e: Seven potential signature lncRNAs selected through univariate COX regression. \u003cstrong\u003eD\u003c/strong\u003e: Coefficient path diagram from LASSO regression. \u003cstrong\u003eE\u003c/strong\u003e: Likelihood ratio test plot for the model from LASSO regression.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/80999b3a734af841da7811af.png"},{"id":90383798,"identity":"dca96b2d-46c0-442f-9af8-4eb430260610","added_by":"auto","created_at":"2025-09-02 07:04:44","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1318617,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation analysis of model lncRNAs. \u003cstrong\u003eA:\u003c/strong\u003e The optimal model combinations established by 5-lncRNAs through multivariate COX analysis. \u003cstrong\u003eB:\u003c/strong\u003e Heatmap of target genes significantly associated with model lncRNAs. \u003cstrong\u003eC:\u003c/strong\u003e Expression levels of model lncRNAs in normal and tumor samples. \u003cstrong\u003eD:\u003c/strong\u003e Expression levels of model lncRNAs in risk subgroups.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/69c4fdfabec2362ad3f5c155.png"},{"id":90381574,"identity":"1c843882-21b6-4f07-af95-cfb549f9b9e4","added_by":"auto","created_at":"2025-09-02 06:48:44","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2185765,"visible":true,"origin":"","legend":"\u003cp\u003eSurvival analysis of internal cohort samples. Survival curves for training cohort and validation cohort samples (\u003cstrong\u003eA\u003c/strong\u003e and \u003cstrong\u003eB\u003c/strong\u003e); Distribution of survival status for training cohort and validation cohort (\u003cstrong\u003eC\u003c/strong\u003eand \u003cstrong\u003eD\u003c/strong\u003e). The ROC curve distribution for training cohort and validation cohort samples (\u003cstrong\u003eE\u003c/strong\u003e and\u003cstrong\u003e F\u003c/strong\u003e). The ROC curve distribution for clinical characteristics of training cohort and validation cohort (\u003cstrong\u003eG\u003c/strong\u003e and \u003cstrong\u003eH\u003c/strong\u003e).\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/b3bf45e3e8721a93b0113ff8.png"},{"id":90383365,"identity":"aa52922d-1341-44a3-a837-aab9d79cff6e","added_by":"auto","created_at":"2025-09-02 06:56:44","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1532796,"visible":true,"origin":"","legend":"\u003cp\u003eSurvival analysis curves for different clinical groups. Risk subgroup survival curves constructed based on clinical characteristics such as patient age (\u003cstrong\u003eA\u003c/strong\u003e and \u003cstrong\u003eB\u003c/strong\u003e), gender (\u003cstrong\u003eC\u003c/strong\u003e and \u003cstrong\u003eD\u003c/strong\u003e), tumor stage (\u003cstrong\u003eE\u003c/strong\u003e and \u003cstrong\u003eF\u003c/strong\u003e), tumor grade (\u003cstrong\u003eG\u003c/strong\u003e and\u003cstrong\u003e H\u003c/strong\u003e), T staging (\u003cstrong\u003eI\u003c/strong\u003e and \u003cstrong\u003eJ\u003c/strong\u003e), and N staging (\u003cstrong\u003eK\u003c/strong\u003e and \u003cstrong\u003eL\u003c/strong\u003e).\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/e3ffaac2c4ef58e54b3523f1.png"},{"id":90381611,"identity":"8739ed47-f135-4828-860b-fbcade17a579","added_by":"auto","created_at":"2025-09-02 06:48:45","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2253911,"visible":true,"origin":"","legend":"\u003cp\u003eSurvival analysis of samples from different databases. Survival curves of samples from TCGA and ICGC database (\u003cstrong\u003eA\u003c/strong\u003e and \u003cstrong\u003eB\u003c/strong\u003e); Distribution of survival status for samples from TCGA and ICGC database (\u003cstrong\u003eC\u003c/strong\u003eand \u003cstrong\u003eD\u003c/strong\u003e). ROC curve distribution for samples from TCGA and ICGC database (\u003cstrong\u003eE\u003c/strong\u003eand \u003cstrong\u003eF\u003c/strong\u003e). ROC curve distribution of clinical characteristics for samples from TCGA and ICGC database (\u003cstrong\u003eG\u003c/strong\u003e and \u003cstrong\u003eH\u003c/strong\u003e).\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/3bfc2a361ebff709ffb93f45.png"},{"id":90383374,"identity":"60621674-a29b-4828-84ed-c58ac2a6e7c1","added_by":"auto","created_at":"2025-09-02 06:56:45","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":775073,"visible":true,"origin":"","legend":"\u003cp\u003eAssessment of clinical characteristics from samples in different databases. Univariate COX regression analysis of clinical characteristics from samples in the TCGA and ICGC database (\u003cstrong\u003eA \u003c/strong\u003eand \u003cstrong\u003eB\u003c/strong\u003e); multivariate COX regression analysis of clinical characteristics from samples in the TCGA and ICGC database (\u003cstrong\u003eC \u003c/strong\u003eand \u003cstrong\u003eD\u003c/strong\u003e).\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/6c58d34f2434b2eadecab38b.png"},{"id":90381602,"identity":"2995ded6-3e37-47f7-acde-0e416e43a8d9","added_by":"auto","created_at":"2025-09-02 06:48:45","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1080791,"visible":true,"origin":"","legend":"\u003cp\u003eSurvival analysis of risk subgroups associated with mutation data. \u003cstrong\u003eA\u003c/strong\u003e: Comparison of TMB differences in risk subgroups; \u003cstrong\u003eB\u003c/strong\u003e: Correlation analysis between risk scores and TMB value. \u003cstrong\u003eC\u003c/strong\u003e: Survival analysis curve for mutation subgroups. \u003cstrong\u003eD:\u003c/strong\u003e Survival analysis curves for samples in the mutation and risk subgroup.\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/58dba95975a48173560ee779.png"},{"id":90381594,"identity":"66303b19-acec-4b13-9a82-edd34c135762","added_by":"auto","created_at":"2025-09-02 06:48:44","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":567888,"visible":true,"origin":"","legend":"\u003cp\u003eThe model lncRNA expression levels in different pancreatic cell lines. The expression of model lncRNA \u003cem\u003eAL157392.3\u003c/em\u003e,\u003cem\u003e BX284668.2\u003c/em\u003e, \u003cem\u003eFAM83A-AS1\u003c/em\u003e, \u003cem\u003eLINC02257\u003c/em\u003e, and \u003cem\u003eTMEM105\u003c/em\u003e in different pancreatic cell lines.\u003c/p\u003e","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/5c60dc35a500da8ccac27d5c.png"},{"id":92091621,"identity":"18718118-77d0-43e2-8ff8-ae53d3c8eb89","added_by":"auto","created_at":"2025-09-24 13:47:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":14977175,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/407612a7-300a-4183-81f9-dd9e6bffc30b.pdf"},{"id":90383367,"identity":"4bb83675-8f13-42f7-b61b-eebd734d5384","added_by":"auto","created_at":"2025-09-02 06:56:44","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":6929093,"visible":true,"origin":"","legend":"","description":"","filename":"1.SupplementaryMaterials1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/145556fca31bdec49956068b.xlsx"},{"id":90381579,"identity":"eded90cf-de73-4dc3-ba65-d4d8cd8186d8","added_by":"auto","created_at":"2025-09-02 06:48:44","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":823193,"visible":true,"origin":"","legend":"","description":"","filename":"2.SupplementaryMaterials2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/adf95573665077d688aa6c35.xlsx"},{"id":90381577,"identity":"ef0309a6-fc3c-4351-b5d7-2be215f3822b","added_by":"auto","created_at":"2025-09-02 06:48:44","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":34593,"visible":true,"origin":"","legend":"","description":"","filename":"3.SupplementaryMaterials3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/f7b6c2d2428f4d5945c1cbdd.xlsx"},{"id":90383362,"identity":"e3cbd040-814c-4cd2-b0de-ddc2171667c1","added_by":"auto","created_at":"2025-09-02 06:56:44","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":99343,"visible":true,"origin":"","legend":"","description":"","filename":"4.SupplementaryMaterials4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/ca924438e153d8e17b5351e0.xlsx"},{"id":90381581,"identity":"94224600-29ce-4531-89cd-31dcd861e0fd","added_by":"auto","created_at":"2025-09-02 06:48:44","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":98483,"visible":true,"origin":"","legend":"","description":"","filename":"5.SupplementaryMaterials5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/71e0ae66cd167bbf7eb0a4aa.xlsx"},{"id":90383364,"identity":"abf44106-5e52-4aec-8bb1-7b12650fff96","added_by":"auto","created_at":"2025-09-02 06:56:44","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":28147,"visible":true,"origin":"","legend":"","description":"","filename":"6.SupplementaryMaterials6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/1fb40bed3e54d91ad1233f42.xlsx"},{"id":90381586,"identity":"995fb9ae-935d-465e-bf95-d2035ad688a2","added_by":"auto","created_at":"2025-09-02 06:48:44","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":17966,"visible":true,"origin":"","legend":"","description":"","filename":"7.SupplementaryMaterials7.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/eac6788c6644d690a077dec6.xlsx"},{"id":90384727,"identity":"4fb37069-94de-480d-b00f-77a1a6511e50","added_by":"auto","created_at":"2025-09-02 07:12:44","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":12913,"visible":true,"origin":"","legend":"","description":"","filename":"8.SupplementaryMaterials8.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/52b7b328f75f5d67646752f2.xlsx"},{"id":90381599,"identity":"eb24a559-1a7a-43e1-8761-df0ea038407d","added_by":"auto","created_at":"2025-09-02 06:48:45","extension":"xlsx","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":10365,"visible":true,"origin":"","legend":"","description":"","filename":"9.SupplementaryMaterials9.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/4e63d30a018b08a4f34ca863.xlsx"},{"id":90381591,"identity":"a614dbbd-812e-4839-89ab-93811ac3a3b9","added_by":"auto","created_at":"2025-09-02 06:48:44","extension":"xlsx","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":15670,"visible":true,"origin":"","legend":"","description":"","filename":"10.SupplementaryMaterials10.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7246386/v1/d12a423b1559e9b2831cd9d6.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Integrating Liquid-Liquid Phase Separation and Tumor Microenvironment Regulation: A lncRNA-Based Prognostic Model for Pancreatic Cancer","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePC has been increasing in both incidence and mortality rates each year. It has been reported that PC ranks 12th in incidence among all cancers and 6th in cumulative mortality rates.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e It is project that by 2030, PC will become the second leading cause of cancer-related deaths in the United States.\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e Studies indicate that the lifetime risk of developing PC worldwide is up to 0.89% (95%CI: 0.88–0.89), and the median overall survival (OS) among all stages of the disease is only 4 months.\u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e–\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e This severe treatment burden imposes a heavy burden on socioeconomic development, making it urgent to explore potential prognostic biomarkers to guide treatment for PC.\u003c/p\u003e\u003cp\u003ePrevious studies have shown that the assessment of clinical risk factors for PC prognosis is unsatisfactory. Known risk factors do not sufficiently explain the development of PC.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e However, in recent years, lncRNA has garnered widespread attention in oncology research.\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e lncRNA influences tumor occurrence and progression through various mechanisms, including gene expression, post-transcriptional regulation, and cell proliferation.\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e Particularly in PC, abnormal expression of lncRNA is closely associated with tumor aggressiveness, metastasis, and prognosis.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e Additionally, the characteristics of the PC microenvironment are distinct, including hypoxia and high fibrosis.\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e lncRNA participates in the construction of this network by regulating immune cells and cancer-associated fibroblasts within the TME.\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e Notably, lncRNA can also influence the expression of the TME by regulating the extracellular matrix, metabolic status, and intercellular signaling.\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eIt is worth noting that the close relationship between lncRNA and LLPS during tumor progression is one of the important factors affecting the TME.\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e lncRNA can form condensates in the nucleus through LLPS, and these condensates not only participate in the regulation of gene transcription but may also act as signal integrators to respond to changes in the microenvironment.\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e Normal biomolecular condensates ensure fundamental cellular functions, while their anomalous forms can lead to cellular dysfunction and potential tumor.\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e Studies have demonstrated that LLPS plays a crucial role in regulating tumor proliferation and metastasis.\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e A significant number of lncRNAs are localized on chromatin and often form RNA-clouds within specific nuclear regions to regulate gene expression. As major components of various membraneless structures, such as nucleolus and paraspeckles, lncRNAs are frequently imbalanced in cancer.\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e Recent research has shown that certain lncRNA subdomains can selectively combine to proteins like NONO/SFPQ, which dynamically oligomerize and recruit additional proteins through LLPS.\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e This process not only participates in gene expression regulation, RNA processing, and cellular stress responses but is also highly sensitive to cellular states and environmental signals.\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e In summary, this study will anchor biomarkers at the mechanistic level of tumors, screening lncRNA based on the LLPS and TME gene sets to give a reference for the prognosis of PC.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cb\u003eData acquisition and integration\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe data included in this study originates from the TCGA and ICGC datasets on the UCSC platform. 165 tumor samples and 171 normal tissue samples originate from the GTEx-TCGA dataset. This dataset is employed for model construction and serves as an internal validation dataset. To enhance the stability of the model, Researchers also include 90 pancreatic tumor samples sourced from the ICGC dataset, which serve as an external validation dataset. Additionally, differential expression analysis of the GTEx-TCGA gene matrix is conducted with the R language software version 4.2.3 (log\u003csub\u003e2\u003c/sub\u003eFC \u0026gt; 1.0, FDR \u0026lt; 0.05, \u003cem\u003eP-\u003c/em\u003evalue \u0026lt; 0.05). Overall, a total of 5,901 genes and 200 lncRNAs are identified as significantly differentially expressed.\u003c/p\u003e\u003cp\u003e\u003cb\u003eConfirm the theme-related target gene and potential lncRNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003eBased on the GeneCards search engine, gene sets are retrieved with the keywords 'TME' and 'LLPS'. Following this, the intersection of these two gene sets is constructed. This intersected gene set is then combined with the 5,901 differentially expressed genes to identify those that are associated with both themes. Subsequently, the limma R package is applied to perform a correlation analysis between the identified lncRNAs and the intersected genes. As a result, 138 lncRNAs that are linked to both 'TME' and 'LLPS' are confirmed as potential lncRNAs for further analysis.\u003c/p\u003e\u003cp\u003e\u003cb\u003eModel construction and cohort grouping\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo test the effectiveness of the model lncRNA, 165 tumor samples from the GTEx-TCGA dataset are randomly divided into two independent cohorts: a training cohort and a validation cohort. The optimal model is constructed with COX regression, and LASSO regression. To assess the performance of the model lncRNA and reduce the risk of overfitting, 10-fold cross-validation is conducted along with 1,000 random perturbations applied to the training cohort. The formula is then applied to calculate the risk score, where lncRNA\u003csup\u003ek\u003c/sup\u003e= coef(lncRNA\u003csup\u003ek\u003c/sup\u003e )× expr(lncRNA\u003csup\u003ek\u003c/sup\u003e ), with \"coef\" representing the survival coefficient and \"expr\" indicating the gene expression level. Based on the median risk score, both the training and validation cohort, as well as the external validation dataset, are all classified into low-risk and high-risk subgroups for further analysis.\u003c/p\u003e\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\text{r}\\text{i}\\text{s}\\text{k}\\:\\text{s}\\text{c}\\text{o}\\text{r}\\text{e}={\\sum\\:}_{\\text{k}=1}^{\\text{n}}\\left({\\text{l}\\text{n}\\text{c}\\text{R}\\text{N}\\text{A}}^{\\text{k}}\\right)\\:$$\u003c/div\u003e\u003c/div\u003e\u003cp\u003e\u003cb\u003eAssess the survival prediction efficacy of model lncRNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe researchers conduct internal validation by comparing survival status and survival curves of risk subgroups within the training and validation cohorts. To further evaluate the predictive efficacy of model lncRNA in the internal dataset, time-dependent ROC curves and clinical characteristics ROC curves are employed. Building on this evaluation, survival analyses and time-dependent ROC curves for the risk subgroups are performed on the entire TCGA dataset. These analyses are also conducted in an external dataset to illustrate the efficacy of the model lncRNA. Furthermore, researchers compare survival differences in clinical characteristics, such as age, gender, tumor grade, and tumor stage, between low-risk and high-risk subgroups. Finally, the reliability of risk scores is confirmed through the construction of univariate and multivariate COX regression analysis, which develops the predictive capability of the model lncRNA.\u003c/p\u003e\u003cp\u003e\u003cb\u003eConstruction of mutation data analysis\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo further illustrate the efficacy of the model lncRNA, this study extracts mutation data from samples in the TCGA database. Researchers perform correlation analyses to reveal differences in mutation data between the low-risk and high-risk subgroups. Following this, the study categorizes the samples into two groups based on the median tumor mutation burden (TMB) coefficient: low-TMB (L-TMB) and high-TMB (H-TMB). Subsequently, researchers construct survival analysis curves for the mutation subgroups. Finally, researchers correlate the mutation subgroups with the risk subgroups to conduct survival analysis. This analysis strengthens the evidence for the model lncRNAs effectiveness and stability.\u003c/p\u003e\u003cp\u003e\u003cb\u003eModel lncRNAs expression validation in cell lines\u003c/b\u003e\u003c/p\u003e\u003cp\u003eIn this study, four cell lines—hTERT-HPNE (HPDE), PANC-1, ASPC-1, and MIA-PACA-2—are selected for qPCR experiments. The primer sequences applied in the research can be found in supplementary materials. Each group of experiments is conducted in triplicate, and the results are averaged, with the standard deviation calculated to assess variability. The qPCR results are then statistically analyzed employing the 2\u003csup\u003e−ΔΔCT\u003c/sup\u003e method and one-way chi-square analysis. Finally, the results are visualized with Prism software version 9.5.1.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cb\u003eSelection of target gene and model lncRNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe research flowchart for this study is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Through the analysis of sample data from the GTEx-TCGA dataset, researchers discover 3469 genes that are differentially expressed \u003cb\u003e(Supplementary Material-1)\u003c/b\u003e. Researchers then obtained 728 theme genes from the GeneCards website \u003cb\u003e(Supplementary Material-2)\u003c/b\u003e. After intersecting the differentially expressed genes with the theme genes, researchers discover 72 target genes \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA\u003cb\u003e)\u003c/b\u003e. From the correlation analysis results, researchers discover 138 lncRNAs linked to target genes, prompting the development of a heatmap showing the top 30 lncRNAs \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB \u003cb\u003eand Supplementary Material-3)\u003c/b\u003e. The training and validation cohorts are validated as independent datasets based on the chi-square method \u003cb\u003e(\u003c/b\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e. Analyzing the sample data from the training cohort, researchers initially identify 7 lncRNAs through univariate COX regression analysis \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC\u003cb\u003e)\u003c/b\u003e. To validate the result, researchers subsequently perform LASSO analysis, conducting 1,000 random cycles \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE\u003cb\u003e)\u003c/b\u003e. Finally, through multivariate COX regression analysis, 5 lncRNAs are selected as model lncRNAs \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA\u003cb\u003e)\u003c/b\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eBasic information of the random cohort.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eClinical Indicators\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eClassification\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTotal\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eValidation\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eTraining\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e-value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge (Year)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026lt;=65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e41(51.25%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e43(53.09%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9399\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e77\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e39(48.75%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e38(46.91%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGender\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e36(45.00%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e38(46.91%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9319\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e87\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e44(55.00%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e43(53.09%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGrade\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e11(13.75%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e14(17.28%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.4424\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e89\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e48(60.00%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e41(50.62%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e46\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e20(25.00%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e26(32.10%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1(1.25%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0(0%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eⅠ\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e10(12.50%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e7(8.64%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.6011\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eⅡ\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e137\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e68(85.00%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e69(85.19%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eⅢ\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1(1.25%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2(2.47%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eⅣ\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1(1.25%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3(3.70%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eT staging\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eT\u003csub\u003e1\u003c/sub\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5(6.25%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1(1.23%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.3240\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eT\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e19\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e8(10.00%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e11(13.58%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eT\u003csub\u003e3\u003c/sub\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e133\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e66(82.50%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e67(82.72%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eT\u003csub\u003e4\u003c/sub\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1(1.25%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2(2.47%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN staging\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eN\u003csub\u003e0\u003c/sub\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e46\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e24(30.00%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e22(27.16%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.8225\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eN\u003csub\u003e1\u003c/sub\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e115\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e56(70.00%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e59(72.84%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eEstablishment and expression differential analysis of model lncRNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe correlation analysis based on the model lncRNAs and differently expressed genes show that the model lncRNAs are related to many co-related genes in LLPS and TMU \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e. Additionally, researchers assess the expression differences of model lncRNAs between tumor and normal samples, as well as within the subgroups of high-risk and low-risk \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD\u003cb\u003e)\u003c/b\u003e. The results indicate significant differences in the expression of model lncRNAs among samples. Furthermore, the results of model RNA are consistent with the multivariate COX result, with \u003cem\u003eAL157392.3\u003c/em\u003e and \u003cem\u003eBX284668.2\u003c/em\u003e as low-risk genes, while the other model lncRNAs are high-risk genes.\u003c/p\u003e\u003cp\u003e\u003cb\u003eInternal dataset survival analysis\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe analysis of risk subgroups reveals significant differences in survival outcomes in both the training and validation cohorts \u003cb\u003e(Supplementary Materials-4,5)\u003c/b\u003e. Specifically, the prognosis for the low-risk subgroup is found to be significantly longer than that for the high-risk subgroup \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e. In both cohorts, the assessment of risk status indicates that the number of patients deaths clearly increases with the risk scores \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD\u003cb\u003e)\u003c/b\u003e. The area under the time-dependent ROC curve (AUC) for both datasets surpass 0.66, effectively demonstrating the prediction capability of model lncRNAs \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF\u003cb\u003e)\u003c/b\u003e. Furthermore, ROC curves for clinical characteristics indicate that the risk scores is superior to other clinical characteristics as an indicator of prognosis (AUC\u0026thinsp;\u0026gt;\u0026thinsp;0.7) \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eG and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eH\u003cb\u003e)\u003c/b\u003e. To further evaluate the predictive performance of the model, this study assesses the survival time in risk subgroups across various clinical categories (age, sex, grade, and stage, etc). The results highlight significant differences between the low-risk and high-risk subgroups in most clinical categories, thereby further confirming the predictive accuracy of the model lncRNAs \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA-L\u003cb\u003e)\u003c/b\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eExternal database validation\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThis study further analyzes the differences in risk subgroups among samples from the TCGA and ICGC databases \u003cb\u003e(Supplementary Materials-6,7)\u003c/b\u003e. Survival curves indicate that the high-risk subgroup has a shorter survival time compared to the low-risk subgroup \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e. Additionally, the assessment of survival status clearly demonstrates that there are fewer patients surviving in the high-risk subgroup \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD\u003cb\u003e)\u003c/b\u003e. The AUCs based on samples from both databases are basically greater than 0.6 \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eE and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eF\u003cb\u003e)\u003c/b\u003e. While the time-dependent ROC curve within the ICGC dataset shows the lowest value of 0.597 during the first year, it is noteworthy that the predictive accuracy of the model exhibits an overall upward trend over time. Moreover, the AUC for clinical characteristics based on samples from both datasets exceeds 0.68, further confirming the predictive efficacy of the model lncRNAs \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eG and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eH\u003cb\u003e)\u003c/b\u003e. Through verification analysis across different databases, this study evaluates the survival prediction capability of the model lncRNAs from multiple perspectives.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eAssessment of clinical features of model lncRNAs\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo further validate the predictive efficacy of the model lncRNAs for prognosis, this study conducts univariate and multivariate COX regression analyses employing clinical characteristics from samples in the TCGA and ICGC datasets. Both univariate and multivariate COX analyses provide consistent results, indicating that the risk score is significantly associated with patient prognosis \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA-D\u003cb\u003e)\u003c/b\u003e. This demonstrates that the risk score can serve as an independent risk factor for clinical prognosis assessment, thereby further highlighting the stability of the model lncRNAs in evaluating prognostic efficacy. Additionally, the analysis indicates that tumor stage in ICGC dataset and tumor N staging in TCGA dataset also serve as independent factors.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eMutation differences in risk subgroups\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo further compare the differences between the risk subgroups, this study analyzes mutation data of the samples \u003cb\u003e(Supplementary Material-8)\u003c/b\u003e. Researchers obtain TMB values for the samples from the TCGA database. The analysis indicates a statistically significant difference in TMB values between the two risk subgroups, demonstrating that mutations vary between them (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA). Additionally, the correlation analysis result demonstrates a significant positive correlation between risk scores and TMB values \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e. Specifically, this analysis shows that as the risk score increases, the TMB value also rises significantly. Furthermore, survival analysis results reveal clear differences in survival time in the subgroups \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eC and \u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eD\u003cb\u003e)\u003c/b\u003e. This finding further highlights the sensitivity and stability of risk genes in predicting the prognosis status of the samples.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eqPCR validation of model lncRNAs expression\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe study analyzes the expression levels of model lncRNAs among different cell lines \u003cb\u003e(Supplementary Materials-9, 10)\u003c/b\u003e. Specifically, the gene expression levels of the HPDE normal pancreatic ductal tissue cell line are selected as a reference. The protective genes \u003cem\u003eAL157392.3\u003c/em\u003e and \u003cem\u003eBX284668.2\u003c/em\u003e demonstrate significantly higher expression levels in the HPDE cell line compared to three other tumor cell lines \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eA and \u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e. This result is consistent with the results of the previous analysis. Furthermore, the high risk genes (\u003cem\u003eFAM83A-AS1\u003c/em\u003e, \u003cem\u003eLINC02257\u003c/em\u003e, and \u003cem\u003eTMEM105\u003c/em\u003e) also exhibit higher expression levels in the three tumor cell lines \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eC-E). In summary, the experimental results from qPCR are consistent with database analyses, which further confirms the accuracy and stability of the model lncRNAs at the biological level.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003ePrevious studies have shown that LLPS has a significant impact on the formation and regulation of lncRNA, and there is a mutual relationship between TME and LLPS.\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e Additionally, the regulatory role of lncRNAs in various cancers has been well confirmed.\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e Building on these insights, this study selects LLPS and TME as focal points, with lncRNAs as the research object, to explore its potential in predicting the prognosis of PC. Given that this study aims to explore the impact of lncRNAs expression related to TME and LLPS genes on the prognosis of PC patients. To ensure the validity of our results, samples with OS of fewer than 30 days are excluded from the analysis (may result from complications such as infections or surgical issues).\u003c/p\u003e\u003cp\u003eThis study employs a combined analysis of COX regression and LASSO regression to select an ideal lncRNA combination. Based on the analysis results, the optimal combination is established, which includes six lncRNAs: \u003cem\u003eAL157392.3\u003c/em\u003e, \u003cem\u003eBX284668.2\u003c/em\u003e, \u003cem\u003eFAM83A-AS1\u003c/em\u003e, \u003cem\u003eLINC02257\u003c/em\u003e, and \u003cem\u003eTMEM105\u003c/em\u003e. Notably, \u003cem\u003eAL157392.3\u003c/em\u003e and \u003cem\u003eBX284668.2\u003c/em\u003e in this combination act as protective genes, and their high expression is beneficial for patient prognosis. Furthermore, the validation of results from risk subgroup survival analyses across multiple databases confirms the superiority of this model. However, researchers observe no statistically significant differences in the Kaplan-Meier survival curves for clinical subgroups in the stage \u003csub\u003eIII\u0026minus;IV\u003c/sub\u003e group. This lack of significance may arise from the limited sample size in this subgroup.\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e Additionally, the time-dependent and clinical characteristics ROC curves for risk subgroups further confirm the superiority of the model lncRNAs. Researchers also note that the superiority of the risk scores is diluted during the evaluation of data from different databases. It is thought by researchers that this finding may be linked to the small sample size in the ICGC dataset. Moreover, it is worth noting that analysis results indicate that the model lncRNAs can serve as independent risk factors for patient prognosis. If sufficient samples become available in the future, the research team will further validate the results. Additionally, the analysis related to TMB values indicates that the differences in survival times among samples within the independent mutation subgroup are not statistically significant. This implies that variations in TMB values have no significant impact on the survival time of these samples. However, the combined analysis of mutation subgroups with risk subgroups reveals significant statistical differences, further highlighting the validity and accuracy of the model lncRNAs. Furthermore, the expression levels of the model lncRNAs across the four cell line groups are consistent with analysis results, providing biological evidence that supports the reliability of the results. The lncRNAs \u003cem\u003eAL157392.3\u003c/em\u003e, \u003cem\u003eFAM83A-AS1\u003c/em\u003e, \u003cem\u003eLINC02257\u003c/em\u003e, and \u003cem\u003eTMEM105\u003c/em\u003e have been previously shown to have a positive correlation with the progression of various cancers, which aligns with the results of this study.\u003csup\u003e\u003cspan additionalcitationids=\"CR35 CR36\" citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eLimitation\u003c/b\u003e\u003c/p\u003e\u003cp\u003eIt is worth noting that, although the study validates the superiority of model lncRNAs from various angles, there are still several limitations inherent in the research. Undeniably, this study is retrospective in nature, which means that its results may be influenced by the intrinsic design of this research method.\u003csup\u003e\u003cspan additionalcitationids=\"CR39\" citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e To enhance the evaluation of model lncRNAs performance, researchers attempt to include more samples; however, corresponding lncRNAs expression information have not yet been obtained from gene matrices in existing databases. This is also the next direction for the research team, which will involve collecting more clinical samples to further explore and validate the risk signature. In summary, this study successfully identifies model lncRNAs associated with both LLPS and TME at the database level and biological level to construct a risk signature for predicting prognosis of PC patients.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study associates the LLPS and TME gene sets through multi-database analysis and biological experiments, establishing a potential independent risk factor composed of lncRNAs to predict the prognosis of PC patients.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthical approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe sample data involved in this study are all sourced from public database. The samples obtained from the public databases comply with the Helsinki Declaration and adhere to ethical standards. All uploaded data are publicly accessible and have received consent from the participants.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent 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\u003eThe manuscript is original and has not been published or accepted for publication elsewhere, either in whole or in part.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and material\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePancreatic sample data are obtained from the UCSC platform (https://xenabrowser.net/datapages/); Gene sets are retrieved from the GenCard website (https://www.genecards.org/), and primer sequences are designed from the NCBI database (https://www.ncbi.nlm.nih.gov/).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that this research is carried out without any commercial or financial relationships that could be interpreted as a potential conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by the 2023 Youth Incubation Fund of The Second Hospital of Tianjin Medical University (Number: 2023ydey07).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor\u003c/strong\u003e\u003cstrong\u003e\u0026rsquo;s\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Contributions\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYW, XS: writing; data analysis; and data visualization. CD, YW, ZL: data analysis. CZ, HY: review and editing. HH: review; editing; supervision. All authors have contributed to the publication of the manuscript and have agreed to its submission.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgement\u003c/strong\u003e\u003cstrong\u003es\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGratitude is extended to the UCSC platform for providing the public data cohorts from TCGA and ICGC; to GenCards and NCBI for their valuable search engines; and to all authors for their careful review and constructive feedback on this manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. \u003cem\u003eCA: a Cancer Journal For Clinicians\u003c/em\u003e. 2024;74(3):229-263. doi:10.3322/caac.21834\u003c/li\u003e\n\u003cli\u003eRahib L, Wehner MR, Matrisian LM, Nead KT. Estimated Projection of US Cancer Incidence and Death to 2040. \u003cem\u003eJAMA Netw Open\u003c/em\u003e. 2021;4(4):e214708. doi:10.1001/jamanetworkopen.2021.4708\u003c/li\u003e\n\u003cli\u003eWang S, Zheng R, Li J, et al. Global, regional, and national lifetime risks of developing and dying from gastrointestinal cancers in 185 countries: a population-based systematic analysis of GLOBOCAN. \u003cem\u003eLancet Gastroenterol Hepatol\u003c/em\u003e. 2024;9(3):229-237. doi:10.1016/S2468-1253(23)00366-7\u003c/li\u003e\n\u003cli\u003eSiegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. \u003cem\u003eCA: a Cancer Journal For Clinicians\u003c/em\u003e. 2024;74(1):12-49. doi:10.3322/caac.21820\u003c/li\u003e\n\u003cli\u003eMackay TM, Latenstein AEJ, Augustinus S, et al. Implementation of Best Practices in Pancreatic Cancer Care in the Netherlands: A Stepped-Wedge Randomized Clinical Trial. \u003cem\u003eJAMA Surg\u003c/em\u003e. 2024;159(4):429-437. doi:10.1001/jamasurg.2023.7872\u003c/li\u003e\n\u003cli\u003eKlein AP. Pancreatic cancer epidemiology: understanding the role of lifestyle and inherited risk factors. \u003cem\u003eNat Rev Gastroenterol Hepatol\u003c/em\u003e. 2021;18(7):493-502. doi:10.1038/s41575-021-00457-x\u003c/li\u003e\n\u003cli\u003eHuang J, Lok V, Ngai CH, et al. Worldwide Burden of, Risk Factors for, and Trends in Pancreatic Cancer. \u003cem\u003eGastroenterology\u003c/em\u003e. 2021;160(3):744-754. doi:10.1053/j.gastro.2020.10.007\u003c/li\u003e\n\u003cli\u003eNojima T, Proudfoot NJ. Mechanisms of lncRNA biogenesis as revealed by nascent transcriptomics. \u003cem\u003eNat Rev Mol Cell Biol\u003c/em\u003e. 2022;23(6):389-406. doi:10.1038/s41580-021-00447-6\u003c/li\u003e\n\u003cli\u003eAhmad M, Weiswald L-B, Poulain L, Denoyelle C, Meryet-Figuiere M. Involvement of lncRNAs in cancer cells migration, invasion and metastasis: cytoskeleton and ECM crosstalk. \u003cem\u003eJ Exp Clin Cancer Res\u003c/em\u003e. 2023;42(1):173. doi:10.1186/s13046-023-02741-x\u003c/li\u003e\n\u003cli\u003eTan Y-T, Lin J-F, Li T, Li J-J, Xu R-H, Ju H-Q. LncRNA-mediated posttranslational modifications and reprogramming of energy metabolism in cancer. \u003cem\u003eCancer Commun (Lond)\u003c/em\u003e. 2021;41(2):109-120. doi:10.1002/cac2.12108\u003c/li\u003e\n\u003cli\u003eHashemi M, Moosavi MS, Abed HM, et al. Long non-coding RNA (lncRNA) H19 in human cancer: From proliferation and metastasis to therapy. \u003cem\u003ePharmacol Res\u003c/em\u003e. 2022;184:106418. doi:10.1016/j.phrs.2022.106418\u003c/li\u003e\n\u003cli\u003eMcCabe EM, Rasmussen TP. lncRNA involvement in cancer stem cell function and epithelial-mesenchymal transitions. \u003cem\u003eSemin Cancer Biol\u003c/em\u003e. 2021;75:38-48. doi:10.1016/j.semcancer.2020.12.012\u003c/li\u003e\n\u003cli\u003eTao J, Yang G, Zhou W, et al. Targeting hypoxic tumor microenvironment in pancreatic cancer. \u003cem\u003eJ Hematol Oncol\u003c/em\u003e. 2021;14(1):14. doi:10.1186/s13045-020-01030-w\u003c/li\u003e\n\u003cli\u003ePereira BA, Vennin C, Papanicolaou M, et al. CAF Subpopulations: A New Reservoir of Stromal Targets in Pancreatic Cancer. \u003cem\u003eTrends Cancer\u003c/em\u003e. 2019;5(11):724-741. doi:10.1016/j.trecan.2019.09.010\u003c/li\u003e\n\u003cli\u003eSempere LF, Powell K, Rana J, Brock AA, Schmittgen TD. Role of non-coding RNAs in tumor progression and metastasis in pancreatic cancer. \u003cem\u003eCancer Metastasis Rev\u003c/em\u003e. 2021;40(3):761-776. doi:10.1007/s10555-021-09995-x\u003c/li\u003e\n\u003cli\u003eGong Y, Gong D, Liu S, et al. Deciphering the role of NcRNAs in Pancreatic Cancer immune evasion and drug resistance: a new perspective for targeted therapy. \u003cem\u003eFrontiers In Immunology\u003c/em\u003e. 2024;15:1480572. doi:10.3389/fimmu.2024.1480572\u003c/li\u003e\n\u003cli\u003eShi M, Zhang R, Lyu H, et al. Long non-coding RNAs: Emerging regulators of invasion and metastasis in pancreatic cancer. \u003cem\u003eJ Adv Res\u003c/em\u003e. 2025;doi:10.1016/j.jare.2025.02.001\u003c/li\u003e\n\u003cli\u003eTong X, Tang R, Xu J, et al. Liquid-liquid phase separation in tumor biology. \u003cem\u003eSignal Transduct Target Ther\u003c/em\u003e. 2022;7(1):221. doi:10.1038/s41392-022-01076-x\u003c/li\u003e\n\u003cli\u003eGuo Q, Shi X, Wang X. RNA and liquid-liquid phase separation. \u003cem\u003eNoncoding RNA Res\u003c/em\u003e. 2021;6(2):92-99. doi:10.1016/j.ncrna.2021.04.003\u003c/li\u003e\n\u003cli\u003eChen W, Li Y, Zhou Q, et al. The cancer-associated fibroblast facilitates YAP liquid-liquid phase separation to promote cancer cell stemness in HCC. \u003cem\u003eCell Commun Signal\u003c/em\u003e. 2025;23(1):238. doi:10.1186/s12964-025-02256-2\u003c/li\u003e\n\u003cli\u003eSomasundaram K, Gupta B, Jain N, Jana S. LncRNAs divide and rule: The master regulators of phase separation. \u003cem\u003eFront Genet\u003c/em\u003e. 2022;13:930792. doi:10.3389/fgene.2022.930792\u003c/li\u003e\n\u003cli\u003eRoden C, Gladfelter AS. RNA contributions to the form and function of biomolecular condensates. \u003cem\u003eNat Rev Mol Cell Biol\u003c/em\u003e. 2021;22(3):183-195. doi:10.1038/s41580-020-0264-6\u003c/li\u003e\n\u003cli\u003eMehta S, Zhang J. Liquid-liquid phase separation drives cellular function and dysfunction in cancer. \u003cem\u003eNat Rev Cancer\u003c/em\u003e. 2022;22(4):239-252. doi:10.1038/s41568-022-00444-7\u003c/li\u003e\n\u003cli\u003eZheng L-W, Liu C-C, Yu K-D. Phase separations in oncogenesis, tumor progressions and metastasis: a glance from hallmarks of cancer. \u003cem\u003eJ Hematol Oncol\u003c/em\u003e. 2023;16(1):123. doi:10.1186/s13045-023-01522-5\u003c/li\u003e\n\u003cli\u003eWang Y, Chen L-L. Organization and function of paraspeckles. \u003cem\u003eEssays Biochem\u003c/em\u003e. 2020;64(6):875-882. doi:10.1042/EBC20200010\u003c/li\u003e\n\u003cli\u003eZhao J, Xie W, Yang Z, et al. Identification and characterization of a special type of subnuclear structure: AGGF1-coated paraspeckles. \u003cem\u003eFASEB J\u003c/em\u003e. 2022;36(6):e22366. doi:10.1096/fj.202101690RR\u003c/li\u003e\n\u003cli\u003eRen J, Zhang Z, Zong Z, Zhang L, Zhou F. Emerging Implications of Phase Separation in Cancer. \u003cem\u003eAdv Sci (Weinh)\u003c/em\u003e. 2022;9(31):e2202855. doi:10.1002/advs.202202855\u003c/li\u003e\n\u003cli\u003eYan X, Zhang M, Wang D. Interplay between posttranslational modifications and liquid‒liquid phase separation in tumors. \u003cem\u003eCancer Lett\u003c/em\u003e. 2024;584:216614. doi:10.1016/j.canlet.2024.216614\u003c/li\u003e\n\u003cli\u003eNozawa R-S, Yamamoto T, Takahashi M, et al. Nuclear microenvironment in cancer: Control through liquid-liquid phase separation. \u003cem\u003eCancer Sci\u003c/em\u003e. 2020;111(9):3155-3163. doi:10.1111/cas.14551\u003c/li\u003e\n\u003cli\u003eLiu Y-T, Cao L-Y, Sun Z-J. The emerging roles of liquid-liquid phase separation in tumor immunity. \u003cem\u003eInt Immunopharmacol\u003c/em\u003e. 2024;143(Pt 1):113212. doi:10.1016/j.intimp.2024.113212\u003c/li\u003e\n\u003cli\u003eBin W, Yuan C, Qie Y, Dang S. Long non-coding RNAs and pancreatic cancer: A multifaceted view. \u003cem\u003eBiomed Pharmacother\u003c/em\u003e. 2023;167:115601. doi:10.1016/j.biopha.2023.115601\u003c/li\u003e\n\u003cli\u003eAshrafizadeh M, Rabiee N, Kumar AP, Sethi G, Zarrabi A, Wang Y. Long noncoding RNAs (lncRNAs) in pancreatic cancer progression. \u003cem\u003eDrug Discov Today\u003c/em\u003e. 2022;27(8):2181-2198. doi:10.1016/j.drudis.2022.05.012\u003c/li\u003e\n\u003cli\u003eZhou S, Shen C. Avoiding Definitive Conclusions in Meta-analysis of Heterogeneous Studies With Small Sample Sizes. \u003cem\u003eJAMA Otolaryngol Head Neck Surg\u003c/em\u003e. 2022;148(11):1003-1004. doi:10.1001/jamaoto.2022.2847\u003c/li\u003e\n\u003cli\u003eHo K-H, Huang T-W, Shih C-M, et al. Glycolysis-associated lncRNAs identify a subgroup of cancer patients with poor prognoses and a high-infiltration immune microenvironment. \u003cem\u003eBMC Med\u003c/em\u003e. 2021;19(1):59. doi:10.1186/s12916-021-01925-6\u003c/li\u003e\n\u003cli\u003eWang H, Ding Y, Zhu Q, et al. LncRNA FAM83A-AS1 promotes epithelial-mesenchymal transition of pancreatic cancer cells via Hippo pathway. \u003cem\u003eCell Cycle\u003c/em\u003e. 2023;22(12):1514-1527. doi:10.1080/15384101.2023.2216507\u003c/li\u003e\n\u003cli\u003ePark M-S, Jeong SD, Shin CH, et al. LINC02257 regulates malignant phenotypes of colorectal cancer via interacting with miR-1273g-3p and YB1. \u003cem\u003eCell Death Dis\u003c/em\u003e. 2024;15(12):895. doi:10.1038/s41419-024-07259-4\u003c/li\u003e\n\u003cli\u003eYin Y, Sun Y, Yao H, et al. TMEM105 modulates disulfidptosis and tumor growth in pancreatic cancer via the \u0026beta;-catenin-c-MYC-GLUT1 axis. \u003cem\u003eInt J Biol Sci\u003c/em\u003e. 2025;21(5):1932-1948. doi:10.7150/ijbs.104598\u003c/li\u003e\n\u003cli\u003eRamgopal S, Benedetti J, Cotter JM. Performing a Multicenter Retrospective Study. \u003cem\u003eHosp Pediatr\u003c/em\u003e. 2025;15(2):e77-e82. doi:10.1542/hpeds.2024-008020\u003c/li\u003e\n\u003cli\u003eGreenland S, Senn SJ, Rothman KJ, et al. Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations. \u003cem\u003eEur J Epidemiol\u003c/em\u003e. 2016;31(4):337-350. doi:10.1007/s10654-016-0149-3\u003c/li\u003e\n\u003cli\u003eKaji AH, Schriger D, Green S. Looking through the retrospectoscope: reducing bias in emergency medicine chart review studies. \u003cem\u003eAnn Emerg Med\u003c/em\u003e. 2014;64(3):292-298. doi:10.1016/j.annemergmed.2014.03.025\u003cstrong\u003e\u003cstrong\u003e\u003cstrong\u003e\u003cstrong\u003e\u003c/strong\u003e\u003c/strong\u003e\u003c/strong\u003e\u003c/strong\u003e\u003c/li\u003e\n\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":"pancreatic cancer, LLPS, TME and lncRNA","lastPublishedDoi":"10.21203/rs.3.rs-7246386/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7246386/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eIntroduction:\u003c/h2\u003e\u003cp\u003eLiquid-Liquid Phase Separation (LLPS), tumor microenvironment (TME), and long non-coding RNA (lncRNA) all have varying degrees of influence on the expression regulation of tumors. However, research on the association of these three in pancreatic cancer (PC) still requires further exploration. This study seeks to establish the relationships among these three themes through bioinformatics and to identify biomarkers that can predict the prognosis of PC patients.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eData sets from The Cancer Genome Atlas (TCGA) and International Cancer Genome Consortium (ICGC) are obtained from the UCSC platform. lncRNAs associated with the LLPS and TME gene sets are screened, and model lncRNAs are identified through comprehensive analysis conducted with least absolute shrinkage and selection operator (LASSO) regression and cox proportional hazards (COX) regression. Additionally, the predictive efficacy of the model lncRNAs is validated through multiple databases and cohorts. Furthermore, the expression of the model lncRNAs is validated at a biological level.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eA comprehensive analysis establishes an optimal combination consisting of 5 lncRNAs. The Kaplan\u0026ndash;Meier curves and receiver operating characteristic (ROC) curves for each cohort demonstrates the superiority of the model lncRNAs characteristics. Additionally, the COX regression analysis of clinical characteristics and the analysis of mutation data further indicates the stability of the model lncRNAs. Furthermore, the expression levels of model lncRNAs in cell lines are consistent with the analysis results.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eThe model lncRNAs identified in this study, which are correlated with LLPS and TME, demonstrate significant potential as independent biomarkers for predicting the prognosis of PC patients.\u003c/p\u003e","manuscriptTitle":"Integrating Liquid-Liquid Phase Separation and Tumor Microenvironment Regulation: A lncRNA-Based Prognostic Model for Pancreatic Cancer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-02 06:48:39","doi":"10.21203/rs.3.rs-7246386/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":"98f98887-33a3-4b5b-b855-ae2fe0bcd7d4","owner":[],"postedDate":"September 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-09-24T13:39:30+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-02 06:48:39","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7246386","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7246386","identity":"rs-7246386","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}