Knowledge atlas and emerging trends on ncRNAs

IntroductionOsteosarcoma is the most common primary bone tumor in children and adolescents, which occurs most frequently in the metaphysis of long bones, especially in the distal femur, proximal tibia, and humerus (1). There are approximately 4.4 cases of osteosarcoma per million children reported annually. Moreover, because of recurrence and metastasis, a poor prognosis means a high fatality rate (2, 3). Although comprehensive treatments including surgery, radiotherapy, and neoadjuvant and adjuvant chemotherapy have improved continuously in recent years, the 5-year survival rate of patients with osteosarcoma has no significant advances (4, 5). Unfortunately, a few patients with clinical metastasis at presentation do not survive for longer than 5 years (6, 7). Thus, it is important to seek new therapeutic methods for patients with osteosarcoma.Osteosarcoma is derived from bone-forming mesenchymal cells with aberrant activation of oncogenes and inactivation of tumor suppressor genes (8–11). Apart from alterations of protein-coding genes, dysregulation of non-coding RNAs (ncRNAs) plays an important role in the regulation of several biological processes of osteosarcoma, including microRNAs (miRNAs), long non-coding RNAs (lncRNAs), small interfering RNAs (siRNAs), piwi-interacting RNAs (piRNAs), and circular RNAs (circRNAs) (12–16). For example, circRNAs are particular single-stranded RNA molecules with closed loops created by nonlinear backsplicing between a splice donor and an upstream splice acceptor. CircRNAs have been confirmed to play crucial roles in the regulatory element of the genome, including sponges of miRNAs and direct interactions with RNA-binding proteins (17–19). In addition, many specific miRNAs deregulated in osteosarcoma tumors can be demonstrated to be involved in cell differentiation, proliferation, and metastasis (20, 21). The roles and mechanisms of non-coding RNAs in osteosarcoma are not fully elucidated; however, studies on them have been increasing rapidly and gaining increased attention.In this study, bibliometrics can be used for the quantitative analysis of hotspots in the literature. The main objectives of this study were to summarize the current status of and reveal future developing trends in the knowledge domain of ncRNAs on osteosarcoma. It is more crucial to propose research frontiers and potential hotspots in the near future.Materials and methodsSource of bibliometric data and search strategiesLiterature data on ncRNAs of osteosarcoma were identified and downloaded from the Science Citation Index Expanded (SCI-Expanded) of the Web of Science Core Collection (WoSCC) database. Medical Subject Heading (MeSH) was applied in the search. Language was limited to English; literature types were limited to original article and review. The timespan for data retrieval was 2000 to 2021.The scientific literature was searched with the following search strategy: (osteosarcoma OR ‘osteogenic sarcoma’) AND (‘long non-coding RNA’ OR ‘lncRNA’ OR ‘long ncRNA’ OR ‘ncRNA’ OR ‘non-coding RNA’ OR ‘miRNA’ OR ‘microRNA’ OR ‘miRNAs’ OR ‘circRNA’ OR ‘circular RNA’ OR ‘ceRNA’ OR ‘competing endogenous RNA’ OR ‘piRNA’ OR ‘Piwi-interacting RNA’). All data utilized in this study were downloaded from known databases, and no ethical approval was required.Data export and extractionThe process of data searches was conducted on a separate day and by two authors independently. All collected data were saved as ‘plain text’ with the basic information by using the function of ‘export’ in WoSCC; the basic information in each publication was collected including published date, title, source journal, author, country/region, institution, funding sources, abstract, keywords, number of citations, and Hirsch index (H-index). The Hirsch index (H-index) is used to quantify an individual’s scientific research output and measure his or her citation impact (22).The above process was conducted by two authors to ensure authenticity and avoid duplication, and any disagreements were solved through discussion, if necessary, with the third author. Moreover, journal impact factors (JIFs) were collected from the 2022 Journal Citation Reports. The detailed literature search and selection process are shown in Figure 1.Figure 1 Flow diagram of the literature search and selection process.Bibliometric and statistical analysisTo obtain a more comprehensive analysis, three bibliometric tools, including an R software package (‘bibliometrix’) and two software packages, were used to conduct this research. Firstly, the bibliometrix package in R 4.0.3 was used to convert and analyze automatically the bibliographic information of the included publications and included the distribution of countries/regions, years of publication, and source journals. Furthermore, academic cooperation networks between countries/regions were conducted. Then, VOSviewer 1.6.16 and CiteSpace 5.7.R5 were further used for mapping and visualization of the bibliometric networks of included publications. VOSviewer is one of the frequently and freely available bibliometric tools for constructing and viewing bibliometric maps (23). In this research, VOSviewer was used to construct and visualize the following network maps: network map of co-citation authors and journals, and co-occurrence analysis of keywords. Based on the frequencies of items where they occur together, co-citation and co-occurrence analysis mean that each item was connected with others by links, which could reflect the relationships of items (24). Meanwhile, in the network visualization, each node represents a different item, such as countries/regions, journals, or keywords, and the different colors of nodes indicated different taxonomies or occurrence frequencies. The sizes of nodes represent the number of citations or occurrences, with bigger nodes representing a higher level of citations or occurrences (25). The links between the nodes reflect the correlation between co-citations or co-occurrences of items, and the thickness of the links represents the strength of the links; thus, total link strength (TLS) is used to quantitatively evaluate the links (23, 25). A detailed description of the maps can be found in the software manual (https://www.vosviewer.com/documentation).In addition to VOSviewer software, we also used another bibliometric software, called CiteSpace, which is a Java-based scientometrics research software package that is used to analyze and visualize the hot spots and research frontiers in the scientific literature of a discipline or knowledge domain in a certain period (26, 27). In the present research, CiteSpace was applied to conduct and visualize the research cooperation relationship of authors and institutions; the timeline view map of co-citation references; and references with the strongest citation bursts. All visualization maps constructed by CiteSpace also comprised nodes and links representing different items similar to those of VOSviewer. CiteSpace is able to construct some distinctive types of visualization maps, such as the cluster view map and the timeline view map. For example, the timeline view map, based on an important function of time-slicing, provides information on some research areas within corresponding time periods by mapping the highly cited and pivotal documents (27). In CiteSpace, betweenness centrality (BC) is a key indicator that could identify the relative importance of a node within the networks, and nodes with the highest BC value (≥0.1) are known as hub nodes that are marked with purple rings (28). A more detailed description of the software can be found in the operational manual (http://cluster.ischool.drexel.edu/cchen/citespace/CiteSpaceManual).R software (version 4.03) and Microsoft Excel 2019 were used for descriptive statistical analysis, data fitting, and plotting graphs of publications including the number of publications per year, original countries/regions, authors, journals, institutions, H-index, and funding agencies. The growth rate of publications over time was computed according to the previously described calculation formula (29).ResultsPublication outputs and trendsAfter removing 120 unqualified publications, a total of 3206 publications obtained from the WoSCC including 2960 articles and 246 reviews met the inclusion criteria from 2000 to 2021. Figure 2 presents the specific numbers of annual publications about ncRNAs of osteosarcoma. Moreover, the cumulative publication number rose rapidly in the
Read More