A bibliometric analysis of m6A methylation in viral infection from 2000 to 2022

The data were obtained from the Web of Science Core Collection (WoSCC) in this study. Due to the rapid update of this database, the literature search was performed on 1 December 2022 to avoid bias. The search terms were as follows: [ALL = (‘m6A’ OR ‘m(6)A’ OR ‘N-6-methyladenosine’ OR ‘N6 -methylation’ OR ‘N-6 methylation’) AND ALL = (‘viral infection’ OR ‘virus’ OR ‘viruses’)]. Only original articles and reviews in English were included.

In this study, the comprehensive publishing characteristics were extracted from the WoSCC database, including the number of articles and citations, H-index, year of publication, country, affiliation, author, journal, references, and keywords. Search results for the target subject were recorded in ‘Full Record and Cited References’, while files were saved and renamed in ‘Plain Text’ file format. The initial data preprocessing involved filtering articles based on the specified criteria. Articles of the types ‘article’ and ‘review article’ were selected, and the language parameter was set to ‘English’ to restrict the downloaded literature to these categories. Irrelevant articles not associated with m6A methylation and viral infections were subsequently removed. The processed data was then imported into CiteSpace software, with the previously established ‘input’ and ‘output’ folders placed in their respective locations within the software. To ensure data integrity, duplicates were removed, addressing redundancies across articles, reviews, and conference papers. The cleaned dataset was then exported for further analysis. CiteSpace software and the bibliometric analysis capabilities of the R software package were employed for literature metrics, visual analysis, and content analysis.

Bibliometric and visual analyzes were performed using CiteSpace software (version 6.1.R3) and the R-based Bibliometrix (4.2.1) package. As a bibliometrics analysis software, CiteSpace can analyze the knowledge database, research content, hotspots and frontiers of a specific research field through a visualized network map. In this study, CiteSpace was used to obtained dual-map overlay of citations and visualize keywords, clusters, and partnerships between institutions and countries. Subsequently, the Bibliometrix R package was used to conduct author's keywords analysis, theme evolution, multiple correspondence analysis, and thematic map to obtain trending topics and future development directions in related research fields. GraphPad Prism 9.0 was used to visualize the annual publication output. JMP 14 Pro software was used to draw the gengrahical distribution of article publications.

A total of 320 documents were collected from the Web of Science Core Collection based on the predefined queries. Excluding 1 duplicate record, 319 eligible records were included in this review. As shown in Fig. 1, the research trend was divided into three stages. The first stage with only 11 researches was from 2000 to 2013, when the average output of documents was 1.6 per year. In the second stage, the literature output increased slightly. There were 8 articles in 2016, 21 articles in 2017, and a drop of 18 articles in 2018, and then increased to 35 and 42 articles in 2019 and 2020, respectively. After 2020 was the third stage, in which 96 articles were produced in 2021, which was twice the output in 2020. Although the output in 2022 was slightly lower than that in 2021, the total output was still large, which indicates that m6A is a promising direction in the field of viral infection.

By categorizing literature based on its publication location and subsequently calculating metrics such as the number of publications, proportion, and growth rate for each category, we aim to gain insights into the research output of different countries or regions in a specific academic discipline or field. 319 records were published in 28 countries. The darker color in Fig. 2A indicates that the country has higher publications, while white indicates no publication output. China ranked first with the highest number of 157 articles in this field. The US with 123 articles ranked second, and The UK ranked third with 19 articles. China and the US were far ahead of other countries. The trend of annual publications of the top 10 countries was shown in Fig. 2B. Germany was the first country to publish articles, but the number of papers published globally did not show an upward trend until 2016. As of 2019, the annual output of the US was significantly higher than any other country, and then China overtook other countries to ranked first in annual publication outputs.

Subsequently, a country-by-country collaboration analysis was performed (Fig. 2C). Among all countries, the US had the greatest centrality (0.44), followed by Israel (0.43) and China (0.28), indicating that they have a strong bridge role in this area. It is worth mentioning that despite relatively few published articles, Israel had close cooperation with other countries, especially with regions such as Cyprus, Croatia, Bangladesh and Bulgaria. In this field of research, the US had a higher central status and more cooperation, such as China, Germany, Canada, France and Switzerland.

Analysis of authors collaboration and cited analysis was used to measure the collaborative relationships among authors and the citation relationships among documents. This analysis serves to reflect the developmental trends, research focal points, and academic exchanges within a discipline by exploring the dynamics of collaboration and influence among researchers. These publications involved 328 authors, with an average of 2.65 articles per author. The first seven authors were all from China. Wang X was the highest producer with 17 papers. WANG X was the most cited author with 181 citations, followed by MEYER KD, DOMINISSINI D and LICHINCHI G (Additional file 2: Table S2). Then, the author collaborations were analyzed. Figure 5A showed the top 7 author collaboration groups. WANG X, ZHANG Y, and LI Y and others formed the largest collaborative team, of which WANG J was circled in purple for greater centrality. SIDDIQUI A, KIM D and others made up the second cooperative team, while HE C and LIU Y formed separate groups. Co-occurrence network analysis of institutions reveals collaborative relationships. The map, with a network density of 0.0232, consists of 236 nodes and 642 links (Fig. 5B). The largest node is the Chinese Academy of Sciences, circled in purple, with a centrality of 0.12, indicating that it plays a critical role in this field. Most institutions have collaborated, such as Duke University, Zhejiang University, University of California San Diego, and The University of Chicago (Fig. 5B left). To explore the research topics between institutions, keywords were clustered using the LLR, with ‘m6A’ as the keyword being the largest cluster (Fig. 5B right).

By analyzing cited journals, cited journal analysis can provide important reference information for subject development, research evaluation, academic exchanges, etc. Among the 576 cited journals, 26 were cited over 100 times (Additional file 3: Table S3). Nature was the most cited journal, followed by Cell and Nucleic Acids Research. Of the top 10 cited journals, 8 were in the Q1 region, while Journal of Virology and Journal of Biological Chemistry belonged to Q2. Nature had the highest impact factor (IF) of 69.504, and Cell ranked the second with an IF of 66.850. The dual-map overlay of journals reflects the relationship between the source journal (left) and the target journal (right), as well as the topics covered in the journals (Fig. 6). The colored links represent the relationship between the two journals, with 1 main broad citation path (yellow path). The result showed that journals involving molecular, biology and immunology were always cited by journals of molecular, biology and genetics, with a Z-value of 7.21 and an f-value of 8510. It is worth noting that we subsequently conducted an accessibility analysis to assess whether articles had been published in open-access journals, aiming to analyze the proportion of open-access publications in this particular research category. The results indicated that out of 319 articles, 245 (76.80%) were freely accessible in the database. This finding underscores the accessibility and impact of the literature in question, emphasizing the prevalence of open-access publications within the database.

As the discipline evolves, the usage of keywords or topics in the literature undergoes changes. Theme evolution and thematic maps can reflect the research interests and focal points within a disciplinary field. This provides valuable insights for the analysis of disciplinary development trends while uncovering research hotspots that guide future research directions within the field. The period from 2000 to 2022 was divided into 4 parts as shown in Fig. 7A. Before 2017, the keywords were m6A and epitranscriptome, and then the research on METTL3 and hepatocellular carcinoma related fields began to increase. In 2021, viral replication and innate immunity of viruses began to receive attention. METTL3 and epitranscriptome were still one of the research priorities. In 2022, the research on m6A methylation related to COVID-19 showed an increased trend and occupied an important position. Multiple correspondence analysis was performed to generate a conceptual structure map of author’s keywords (Fig. 7B). Inferred from the results, the keywords were divided into a large category and a small category. The first category marked in red involved most of the important concepts in the research field, which is highly consistent with the study of m6A methylation in viral infection, where m6A, innate immunity, METTL3/14, viral replication and hepatitis virus, etc. were important factors. In addition, prognosis, RNA binding proteins, post-transcriptional gene regulation, etc. have also attracted considerable research interest. Thematic map was implemented to assess the importance and future trends of keywords (Fig. 7C). The first quadrant, representing ‘motor themes’, signifies the research hotspots within the field, serving as focal points for future investigations. These areas exhibit high research density and significant influence, indicating substantial progress in the field and widespread academic attention. The second quadrant, labeled as ‘Niche themes’, indicates a solid research foundation within the field without clear dominance. While these areas demonstrate high research density, their impact is comparatively low, suggesting a cumulative body of research without breakthrough achievements. The third quadrant, denoted as ‘Emerging or declining themes’, reflects research in its early stages, requiring further development. These areas have low research density and low impact, signifying an exploratory phase that necessitates additional experience and accumulation of knowledge. The fourth quadrant, termed ‘Basic themes’, represents research with potential practical applications, warranting further exploration. Although these areas have low research density, their impact is high, indicating promising prospects but still in the nascent stages of development. In this study, the first quadrant (upper right) showed the most important themes, implying both important and well-developed. METTL3, plant viruses, cancer progression and type I interferon (IFN-I) were included, reflecting a good development trend and may be future research hotspots. The third quadrant (lower left) showed emerging or declining themes. post-transcriptational modification and RNA-seq might be defined as marginal topics.