The Science Citation Index-Expanded (SCI-Expanded) of WoSCC was searched for publications related to ncRNAs in PC. The retrieval strategy used in this study was “TS = [(“miRNA*” OR “microRNA*” OR “miR-*” OR “lncRNA*” OR “long noncoding RNA*” OR “long non-coding RNA*” OR “circRNA*” OR “circular RNA*” OR “circ_*” OR “non-coding RNA*” OR “noncoding RNA*”) AND (“prostate cancer” OR “prostate neoplasm”)].” The search time span was set to between 1 January 2004 and 31 December 2023. In the first stage of screening, the type of publication was restricted to Article and the language was limited to English. In the second stage, irrelevant publications were excluded based on the title, keyword, abstract, and full text. The full records and cited references of eligible articles were exported from SCI-Expanded in plain text format. The flow chart of research steps of this study is shown in Figure 1.

All valid data were imported into Origin (version 2021) or CiteSpace (version 6.2. R6) for analysis and visualization. Origin was used to conduct statistical analysis on the annual number of publications and their citations, and to visualize the collaborations between different countries. CiteSpace was performed for co-operation analysis of countries, institutions, and authors. In addition, we used CiteSpace to draw a dual-map overlay of journals to investigate the evolutionary relationship between knowledge topics in directly citing journals and cited journals. In keyword analysis, we merged keywords with the same meaning to get a better perspective. The co-occurrence analysis of keywords and the co-citation analysis of references were performed by CiteSpace. Moreover, burst detection in CiteSpace was used to detect sudden surges in references and keywords at a certain time.

A total of 2,951 articles on ncRNAs in PC were screened from 7,630 records retrieved from WoSCC. According to the number of citations, the total citations of the top 10 most cited papers accounted for 43.87% of all papers (Supplementary Table S1). The most cited article was titled “c-Myc suppression of miR-23a/b enhances mitochondrial glutaminase expression and glutamine metabolism,” which had 1,641 citations in Nature. The article published in Cell titled “The Landscape of Circular RNA in Cancer” had the highest annual citation frequency. The distribution of publications and their citations is presented in Figure 2. From 2004 to 2023, the cumulative number of global publications showed a consistent upward trend. Specifically, the evolution of PC-related ncRNA research consisted of threee stages. The field was in its early stages from 2004 to 2008, when fewer than 20 papers were published each year. The second stage was from 2009 to 2021, with a rapid increase in the number of annual publications. As the third stage from 2022 to 2023, the number of annual publications was in a steady growth, and the cumulative number of publications in 2023 was 113.5 times that of 2008. The cumulative number of citations for all articles in 2023 was 21,846, with an average of 7 citations per article. These results indicate that ncRNA research is an active field in PC and has received extensive attention from scholars.

Analyzing the research output of countries/regions and institutions provides insights into the global distribution and trends within a specific field, enabling the identification of influential countries/regions and institutions. A total of 78 countries/regions were involved in the publications of ncRNAs in PC. Among them, China (1,665), the United States (699), Germany (131), Canada (120), and Italy (112) were the top five productive countries/regions in terms of the number of publications (Table 1). The H-index is a hybrid quantitative metric used to measure academic impact. The United States (119) has the highest H-index, indicating its leading academic influence in PC-related ncRNA research, followed by China (88), Canada (47), Japan (46), and Italy (43). Centrality helps identify key roles in information flow and knowledge dissemination within a network. The United (0.56) States exhibited the highest centrality, followed by China (0.27), the United Kingdom (0.2), Canada (0.11), and Germany (0.1). As depicted in Figure 3A, the United States collaborated with 26 other countries, notably including China, Canada, and the United Kingdom, which underscores its significant role as a central hub in international collaborations.

Based on the number of publications, Table 2 presents the top 10 institutions of ncRNA research in PC, with 9 of them located in China. Fudan University (98) had the highest number of publications, followed by Shanghai Jiao Tong University (96), Nanjing Medical University (91), and Sun Yat-sen University (91). Furthermore, the H-index of Sun Yat-sen University (36) and Harvard University (36) was the highest, followed by Fudan University (34) and Shanghai Jiao Tong University (33). Visualized with CiteSpace, we mapped an institutional collaboration network consisting of 102 nodes and 112 linkages (Figure 3B). Southern Medical University (0.41) had the highest centrality, followed by Guangzhou Medical University (0.36) and Nanjing Medical University (0.29). As the institution with the most publications, Fudan University primarily had close collaborations with Naval Medical University, Tongji University, and Chinese Academy of Sciences.

Author analysis facilitates the identification of pivotal contributors within a field and offers insights to guide academic exchanges and project collaborations. The top 10 active authors are listed in Table 3. Rajvir Dahiya contributed the most research with 34 publications, followed by Sharanjot Saini (29) and Yuichiro Tanaka (29). According to authors’ H-index, Rajvir Dahiya (23) ranked first, followed by Naohiko Seki (22), Sharanjot Saini (20), Yuichiro Tanaka (20), Shahana Majid (20), and Arul M Chinnaiyan (20). Rajvir Dahiya made significant contributions, with his research emphasizing the regulatory roles of miRNAs in the proliferation and metastasis of PC (Noonan et al., 2009; Saini et al., 2011). Co-cited authors are those cited together in one or more publications, used to identify key researchers in a particular academic field. Supplementary Table S2 shows that Rebecca L Siegel (978) had the most co-citations, followed by David P Bartel (451) and Ahmedin Jemal (394). These scholars have established the groundwork for research on ncRNAs in PC.

Journal analysis provides researchers with prioritized options for submitting their work and accessing articles within a specific field. A total of 556 academic journals included articles on PC-related ncRNAs. The top 10 journals included 658 papers in the field, accounting for 22.30% of all publications (Table 4). PLoS One (102) was the most published journal, followed by Prostate (97), Oncotarget (95), Oncogene (61), and Scientific Reports (60). Cancer Research (6,656), PLoS One (6,567), and Oncogene (5,663) were the top three most frequently cited journals.

The top 20 active co-cited journals are presented in Supplementary Table S3. Among them, Cancer Research (5,615) had the most co-citations, followed by Oncogene (3,325), PLoS One (3,190), Cell (3,034), and Proceedings of the National Academy of Sciences (2,720). In addition, to investigate the evolutionary relationship between knowledge topics in directly citing journals and cited journals, we used CiteSpace to draw a dual-map overlay of journals, in which the topics of directly citing journals are distributed on the left and the topics of cited journals are represented on the right. The analysis presented in Supplementary Figure S1 revealed a prominent citation pathway, indicating that research published in Molecular/Biology/Genetics journals was predominantly referenced by studies from Molecular/Biology/Immunology journals.

To track the knowledge structure and research trends in the field of PC-related ncRNAs, we conducted a co-citation analysis of the references from 2,951 articles with CiteSpace. Supplementary Table S4 shows the top 10 frequently co-cited references. Among them, “(Siegel et al., 2017)” was the most co-cited reference, with 406 citations. Subsequently, the clustering analysis of CiteSpace was performed to construct a map of co-cited references, which consisted of 842 filtered nodes, 1917 connections, and 10 clusters (Figure 4). The modularity Q score was 0.7094 (>0.3), and the mean silhouette S was 0.867 (>0.7), demonstrating both a significant clustering structure and a highly homogeneous clustering network. Supplementary Figure S2 displays the timeline view of co-cited references, highlighting the evolution of each cluster over time. The largest cluster was “EVs” (cluster #0), followed by “PC” (cluster #1) and “circRNA” (cluster #2). According to the results of cluster analysis, “microarray” (cluster #7), “PC” (cluster #1), and “HRPC” (cluster #8) represented the early knowledge base in the field of PC-related ncRNAs in the past 20 years, while “EVs” (cluster #0), “circRNA” (cluster #2), and “ceRNA” (cluster #4) reflected the current research direction.

Furthermore, through the burst detection function of CiteSpace, we explored the top 15 co-cited references with strong citation bursts (Table 5). The paper that first exhibited strong citation burst over the past 20 years was “A microRNA expression signature of human solid tumors defines cancer gene targets,” authored by Stefano Volinia and published in 2006. The paper titled “microRNA expression in human prostate cancer” published in 2008 received the strongest citation burst (46.12), followed by “Kati P Porkka, 2007” (42.48), and “Stefan Ambs, 2008” (38.65). These papers constitute significant events in the current knowledge base of PC-related ncRNA research.

Identifying the high-frequency keywords in research on PC-associated ncRNAs facilitates the rapid recognition of the hot topics of interest within the academic community. Supplementary Table S5 lists the top 50 most frequent keywords, with the 15 highest-ranking ones being “PC” (1938), “expression” (1,258), “miRNA” (809), “proliferation” (688), “metastasis” (632), “progression” (521), “growth” (492), “cancer cell” (490), “invasion” (447), “gene” (360), “biomarker” (352), “lncRNA” (329), “androgen receptors” (284), “cancer” (268), and “apoptosis” (265). Among these high-frequency keywords, those pertaining to ncRNAs include “miRNA” (809), “lncRNA” (329), and “circRNA” (82). The co-occurrence analysis of keywords helps elucidate the relationships between different knowledge points in PC-associated ncRNA research, thereby contributing to the construction of a knowledge map for the field. To obtain a better network visualization of the keyword co-occurrence, we merged similar keywords from 2,951 articles, resulting in a co-occurrence map with 129 filtered nodes, 137 connections, and 11 clusters (Figure 5A). The modularity Q score was 0.8081, and the mean silhouette S reached 0.9215, indicating the highly convincing keyword network. In addition, we presented a visualization of the evolution of keywords in different clusters over time in the form of a timeline view (Supplementary Figure S3). “gene” (cluster #0) was the largest cluster, encompassing keywords such as “miRNA,” “mRNA,” “polymorphism,” “amplification,” “receptor.” and “activation.” “progression” (cluster #1) was the second largest cluster, including keywords such as “invasion,” “growth,” “signature,” “circRNA,” and “cancer.” Followed by “lncRNA” (cluster #2), including keywords such as “mechanism,” “ceRNA,” “mesenchymal transition,” “chromatin,” and “sequence.”

Burst detection of keywords is employed to identify new or rapidly increasing keywords, which may represent cutting-edge topics or novel areas of research. By using CiteSpace, we obtained the top 10 keywords with the strongest bursts (Figure 5B). The top three keywords with the strongest burst were “circRNA” (19.52), “gene” (18.64), and “signature” (14.52). Notably, the keywords “circRNA,” “EVs,” and “exosome” continued to show significant burst activity in 2023, indicating that these topics are key directions for future research. In addition, we extracted all miRNAs, lncRNAs, and circRNAs from the keywords, and merged and standardized similar miRNAs. Table 6 lists the frequently occurring miRNAs, lncRNAs, and circRNAs, all of which have been confirmed to play significant roles in the development of PC.

With the rapid development of information technology and increased medical academic papers, bibliometrics has been applied in various medical fields and plays an irreplaceable role in quantitative analysis. In this study, we retrieved publications indexed in WoSCC over the past 20 years and excluded records unrelated to ncRNA research in PC. Ultimately, 2,951 published articles were included in our bibliometric analysis. Between 2009 and 2023, there was a marked increase in both the total number of publications and citations concerning ncRNA research in PC, underscoring its emergence as a rapidly expanding area of interest. China had the highest number of publications and the second-highest number of citations, which shows its significant contribution in this field. As the most influential country, the United States collaborated with 26 countries/regions and maintained its closest partnership with China. In terms of institutional distribution, Fudan University, which had the highest number of publications, primarily collaborated with the Naval Medical University, Tongji University, and Chinese Academy of Sciences. The research focus of these institutions mainly revolves around understanding the mechanisms by which miRNAs (Li et al., 2014; Xiang et al., 2015), lncRNAs (Yao et al., 2020), and circRNAs (Kong et al., 2021) contribute to the progression of PC. According to the author analysis, Rajvir Dahiya was the most active and influential author in this field.

Additionally, we identified the top 10 most cited articles in ncRNA research related to PC. The most cited paper was published in Nature by Gao et al. (2009), which was groundbreaking in linking changes in miR-23a/b expression and glutamine metabolism to PC progression and proposed that c-Myc plays a crucial role in regulating crucial metabolic pathways in PC. The paper with the highest annual citation frequency was published by Vo et al. (2019) in 2019 in Cell, which highlighted the potential of circRNAs as diagnostic biomarkers and therapeutic targets for PC.

We further explored the theoretical foundations of PC-related ncRNA research through co-citation analysis of the references. Through CiteSpace clustering analysis, all co-cited references were categorized into a network map containing 10 clusters with diverse labels, where “EVs” (cluster #0), “circRNA” (cluster #2), and “ceRNA” (cluster #4) reflected the current research trends. The top 10 frequently co-cited references included four epidemiological studies and six foundational studies. These epidemiological findings consistently highlighted PC as a leading cause of cancer-related deaths in men and a significant public health concern in China (Siegel et al., 2014; Chen et al., 2016; Siegel et al., 2017; Sung et al., 2021). Among the six foundational studies, four investigated microRNA profiles in PC (Porkka et al., 2007; Ambs et al., 2008; Ozen et al., 2008; Schaefer et al., 2010), whereas the other two focused on exploring the molecular and genomic characteristics of the disease (Taylor et al., 2010; Network, 2015). Moreover, we identified the top 15 popular references in PC-related ncRNA research based on citation bursts, which include most of the highly co-cited references mentioned above. It is worth noting that the study by Mitchell et al. (2008) underscored the potential of circulating miRNAs in serum or plasma as stable blood-based biomarkers for cancer classification and prognostication, specifically demonstrating that miR-141 effectively distinguished PC patients from healthy controls. This research provided the first strong evidence supporting the diagnostic utility of circulating miRNAs in PC. Ribas et al. (2009) were the first to discover that high expression of miR-21 promoted the growth of PC and conferred castration resistance, while inhibition of miR-21 reduced androgen-driven proliferation of PC cells. This study provided crucial molecular insights into targeting miRNAs for the treatment of PC. In summary, co-citation analysis of the references enables us to better understand the knowledge structure of PC-related ncRNAs and to identify the core references within this field.

To explore the research hotspots in PC-related ncRNAs, we analyzed the keywords from 2,951 articles. High-frequency keywords with co-occurrence rates exceeding 500 included “PC,” “expression,” “miRNA,” “proliferation,” “metastasis,” and “progression,” indicating that research primarily focused on the molecular mechanisms of miRNAs in PC proliferation and metastasis. Commonly mentioned miRNAs included miR-21, miR-200, miR-375, miR-34, miR-145, miR-141, miR-205, miR-1, miR-143, miR-17, and miR-182. Aside from “miRNA,” “lncRNA” and “circRNA” were also high-frequency keywords associated with ncRNAs, highlighting their significant relevance in the field of PC. We also used CiteSpace for keyword burst detection, which revealed that recent emergent keywords include “circRNA,” “EVs,” and “exosome,” all of which represent cutting-edge areas in ncRNA research related to PC. Here, we provide a brief discussion on the potential of circRNAs and EVs in PC.

In this study, we conducted the first bibliometric analysis using data from the WoSCC to objectively assess the research trends and current landscape in the field of PC-related ncRNAs. While our approach offers valuable insights, it is important to recognize certain limitations. First, although WoSCC is widely considered the premier database for bibliometric studies, there remains the possibility that some relevant articles were not captured, potentially introducing minor biases in our findings. Moreover, our analysis was based on data from publications spanning 2004 to 2023, limiting its ability to reflect the most recent advances in the field. Third, the nature of bibliometric analysis tends to prioritize highly cited mainstream research, potentially overlooking less cited but innovative studies. Nevertheless, we believe these limitations do not substantially diminish the relevance or contributions of the publications included in our analysis to the field.