Bibliometric analysis and network visualization on Tuberous Sclerosis Complex

: Background: Tuberous sclerosis complex (TSC) is a rare autosomal dominant multisystem disease resulting from hyperactivation of the mammalian target of rapamycin (mTOR) signaling pathway. This study aimed to measure the quantitative impact of publications in TSC. Materials and methods : We analysed TSC literature obtained from the Scopus database using Bibliometrix R Package and VOSviewer software. Annual publication trends, most productive and collaborative authors/institutions/ countries, most cited articles, most popular journals and author’s keywords were presented using standard bibliometric indicators. Results and discussion : A total of 5375 documents on TSC were published from 1960 to December 2020, with an increasing trend. The three primary contributing writers were Curatolo P, Kwiatkowski DJ, and Thiele EA, with the United States and its institutions being the largest contributor. The research identified two of the most referenced papers as TSC’s seminal pieces. The top journals that published TSC research were medical journals, namely Journal of Child Neurology, Epilepsia , and Pediatric Neurology . mTOR inhibitor, everolimus


Introduction:
Tuberous sclerosis complex (TSC) is an autosomal dominant multisystem disease characterised by tumorigenesis, neurologic and behavioural impairments 1 .It affects both men and women, as well as all ethnic groups 2 .It is a rare disease and affects one in every 6000 to 10,000 newborns annually 3 .TSC1 is located on the long arm of chromosome 9 (9q34), while TSC2 is on the short arm of chromosome 16.(16p13.3).A mutation in TSC1 or TSC2, which encode the proteins hamartin and tuberin, respectively 4 , is the most prevalent cause of TSC 5 .
TSC1 mutations are usually nonsense or frameshift mutations that cause protein truncation, whereas TSC2 mutations are more likely to be missense mutations, substantial deletions, or rearrangements 6 .TSC1 and TSC2 mutations have been discovered in 10-20 per cent and 70-90 per cent of TSC patients, respectively 7,8 .However, no mutation is found in 10-15% of individuals with a clinical diagnosis of TSC 98 .These individuals are thought to have somatic mosaicism, which may explain their symptoms.A third candidate gene could also be involved, or the disease could be caused by an intronic mutation 9 .The research on the genetic basis of TSC is still being studied in order to better understand its clinical symptoms.
Pathogenic TSC2 mutations are associated with a more severe clinical phenotype than mosaic TSC2 or TSC1 variations [10][11][12] .Both TSC1 and TSC2 are tumour suppressor genes that, when turned off, cause uncontrolled cell cycle progression and the spread of hamartomas throughout the body 4,13 .The majority of TSC patients have mutations in TSC1 and TSC2, leading to hyperactivation of the mammalian target of rapamycin (mTOR) signalling pathway and abnormalities in a variety of cell functions 14 .Because TSC protein products of TSC, hamartin and tuberin, act within the same intracellular pathway, their mutations induce almost similar disease phenotypes 15 .While TSC can affect any organ system in the body, certain organs such as the heart, kidneys and eyes are more vulnerable than others 1 .
Bibliometric studies on the mTOR signalling pathway in liver and kidney diseases have recently been published 16,17 .There have, however, been no published bibliometric or visualization studies on TSC research.The purpose of this research is to conduct an in-depth bibliometric analysis and network visualization on TSC literature.The following research questions (RQs) will be addressed: RQ1: What are the general descriptions regarding

Search Strategy and Data Extraction:
A literature search for all the published TSC documents was conducted on 4 July 2021, using the Scopus database.The following search terms in the article title were used: ( "Tuberous sclerosis*" OR "tuberose sclerosis*" OR epiloia OR "Bourneville-Pringle disease" ).A total of 5590 documents were identified throughout the search, however, 215 were excluded including 2021 documents (n = 171) and erratum (n=44).The summary of the search strategy is illustrated in Figure 1.Citation information, bibliographical information, abstract, keywords and other information including references were extracted from each publication and exported in comma-separated values (csv) file format for further analysis.

Data Analysis and Visualization:
The following bibliometric and network analyses were conducted to answer the RQs mentioned earlier.The bibliometric analysis was conducted using R software version 3.6.2(R Foundation for Statistical Computing, Vienna, Austria; http://www.r-project.org) via the Bibliometrix R package 18 .While the network analysis was conducted using VOSviewer software (version 1.6.15).

General description related to TSC documents:
Table 1 presents the main information regarding the retrieved documents.A total of 5375 documents from 1830 sources were retrieved from the Scopus database.All these documents used 11732 keywords plus and 3633 author keywords.The TSC literature we used covered the years 1906 through 2020.These documents were written by 17102 different authors, with just 526 (3.08%) having a single author.The TSC documents have a high collaboration index and only 659 out of 5375 documents (12.26%) were written by a single author.The authors per document were 3.18, which means, on average, more than three authors have written one document.Note: Authors per Document index is calculated as the ratio between the total number of authors and the total number of documents.The Co-Authors per Documents index is calculated as the ratio between the author appearances and the total number of documents.
The majority of the retrieved publications were original articles (n = 4328, 80.5%).Other documents were review articles (n = 371, 6.9%), letters (n = 300, 5.6%), notes (n = 134, 2.5%), conference papers (n = 114, 2.1%), and book chapters (n = 60, 1.1%).To a much lesser extent (n = 68, each ≤ 1%), editorials, short surveys and books were also published.Regarding the subject areas, the majority of the publications were classified under medicine (n = 4861, 70.2%), although several documents were published in Neuroscience (n = 818, 11.8%), and Biochemistry, genetics and molecular (n = 688, 9.9%).To a much lesser extent (n = 554, each <2%), the publications were under Arts and humanities, Multidisciplinary, Dentistry, Psychology, and others.1982, 1985, 1987 (2), 1990, 1991 (3), 1993, 1994, 1995,   1996, 1997, 1998, 1999, 2000, 2001 (2)  6), 2016 (7), 2017 (7), 2018        The network visualization map of country coauthorship produced by VOSviewer is shown in Figure 4.Each node represents a country and the node size is the number of articles published by the country.There were 27 out of 220 countries with at least 20 papers and 100 citations.A total of 4 clusters was identified and most of the productive countries shared the same cluster.For example, United States and Italy belong to the green cluster, Japan, China, India, France and Spain belong to the red cluster, and Germany and Turkey belong to the blue cluster.Notes: a Percentage calculated out of the retrieved 5375 documents.

Most frequent journals:
Table 6 presents the most frequent journals where the 5375 selected documents were published in.Journal of Child Neurology was the most frequent journal with the highest h-index, g-index, m-index and total citations, followed by Epilepsia and Pediatric Neurology.

Most frequent keywords:
Table 7 presents the top 10 frequency of occurrence of the author's keywords and keywords plus.The occurrence of the author's keywords throughout the years is shown in Figure 8.The author keywords "tuberous sclerosis" was prominent from the 1970s until 2012 when "tuberous sclerosis complex" overtook and became the most prominent.
Figure 5 shows an overlay visualisation map of the author's keywords created using VOSviewer.There were 101 out of 3526 keywords with at least 10 co-occurrence.The evolution of the author's keywords revealed that some keywords such as tuberous sclerosis, hamartoma, tuberin, hamartin, Bourneville-pringle disease, mental retardation, computed tomography, phakomatosis and prenatal diagnosis were older (purple), while    In terms of productivity, these journals show an increasing trend.On the other hand, high quality, broad-scope journals such as Lancet, Journal of Medical Genetics and Annals of The New York Academy of Sciences showed decreasing popularity.This could be due to the relative exhaustion of the genetic pathology of the disorders.
The results of the most frequently used keywords provide a clear picture of the current state of TSC research.Essential concepts and knowledge structure of a scientific area could be revealed by examining the connections between keywords in literature 35 .The most frequent author's keywords were TSC, epilepsy, everolimus, angiomyolipoma, mTOR, TSC2, rapamycin, subependymal giant cell astrocytoma, lymphangioleiomyomatosis, TSC1, autism, magnetic resonance imaging, children, renal angiomyolipoma, and seizures.From these keywords, we can deduce that this research field is focused on two primary areas: (i) clinical aspect (children, epilepsy, angiomyolipoma, subependymal giant cell astrocytoma, lymphangioleiomyomatosis, autism, renal angiomyolipoma, seizures and magnetic resonance imaging) and (ii) role of mTOR inhibitors (everolimus and rapamycin) in the treatment of TSC.The evolution of the author's keywords revealed that certain keywords such as mTOR inhibitor, everolimus, sirolimus, mTORC1, mTOR pathway, autophagy, inflammation, infant, intellectual disability, white matter, TSC-associated neuropsychiatric disorders, TuberOus SClerosis Registry to Increase Disease Awareness (TOSCA) and quality of life were relatively new.This suggests that new therapy options targeting vulnerabilities in TSCrelated pathways will continue to be developed and TSC serves as a model for many disorders involving the mTOR pathway 1 .In addition, the TOSCA study

Figure 1
Figure 1The search strategy used in this study

Figure 2
Figure 2 depicts the annual publication trends on TSC research.The first publication on TSC dates back to 1906 and the number of publications was small at the start but increased to 2-digit beginning in 1954 and reached 3-digit in 1991.

Figure 2
Figure 2 Annual publication trendsMost productive authors:Table2presents the top 10 most productive authors.Curatolo P, Kwiatkowski DJ and Thiele FA were the three major contributing authors based on the number of articles each had written.Curatolo P was also the top author of publication over time; he started in 1982 and was actively publishing until 2020.This was followed by Northrup H (1987), Jozwiak S (1990), Kwiatkowski DJ (1991) and Henske EP (1995).

Figure 3
Figure3shows the network visualization map of author co-authorship obtained from VOSviewer.Each node represents an author, the node tag is the author's last name, and the node size is the number of articles published.There were 42 out of 16705 authors who had at least 20 papers and 100 citations.A total of 8 clusters was identified and most of the prolific authors shared the same cluster.For example, Curatolo P, Jozwiak S and Kotulska K belong to the blue cluster, Kwiatkowski DJ and Thiele EA belong to the brown cluster, Sahin M and Northrup H belong to the red cluster, and Franz DN and De Vries PJ belong to the purple cluster.

Figure 4
Figure 4 Co-authorship countries network Most productive institutions: Table 4 presents the top 10 most productive research institutions in publishing on TSC.Of all the 5375 retrieved documents, these institutions published 1802 references (33.5%), of which 1332 (24.5%) were from institutions located in the United States.The rest of the research findings were published by institutions in Poland (n = 147), United Kingdom (n=134), Netherlands (n=100) and China (n = 89), accounting for 8.7% of the total published references.

Figure 5
Figure 5 Overlay visualization of authors' keywords.Discussion: This bibliometric study presents, for the first time, a comprehensive analysis of TSC research using the Scopus database.The majority of the TSC research documents have been published as original articles in the medical subject field.The development of scientific publications from 1906 to 2020 demonstrates that TSC research has increased in relevance over the recent decades, making it an

Table 2
Details of most productive authors

Table 3
States was the most productive country publishing TSC research.It contributed almost a quarter of the scientific production, followed by Japan, the United Kingdom and Italy.Regarding intra-country collaboration, the United States also had the most presents the most productive countries based on the number of published documents.The United

Table 3
Most productive countries Notes: SCP, single country publications; MCP, multiple country publications.a Percentage calculated out of the retrieved 5375 documents.b Multiple country publication ratio was calculated as MCP divided by the total of published documents per country.

Table 4
Most productive institutions

Table 5
presents the most cited documents.The most cited document authored by the European chromosome 16 Tuberous Sclerosis Consortium and published in Cell in 1993 was "Identification and characterization of the tuberous sclerosis gene on chromosome 16".The second most cited document, "Identification of the tuberous sclerosis gene TSC1 on chromosome 9q34", was authored by Van Slegtenhorst M and published in Science in 1997."The tuberous sclerosis complex", authored by Crino PB and published in New England Journal of Medicine in 2006, was the third most cited document.

Table 5
Most cited documents 21Identification

of the tuberous sclerosis complex-2 tumor suppressor gene product tuberin as a target of the phosphoinositide 3-kinase/Akt pathway
24Tuberous

sclerosis complex gene products, tuberin and hamartin, control mTOR signalling by acting as a GTPase-activating protein complex toward Rheb
26Tuberous

Table 7
Most frequent keywords 29,30nd purple (de Vries PJ, Franz DN), as seen in the co-authorship network of authors.The co-authorship networks are prominent bibliometric indicators for illustrating distinct patterns of co-authorship of academic fields29,30.Interestingly, the results of the most prolific journals revealed that TSC research was published in clinical medicine journals, namely Journal of Child Neurology, Epilepsia, Pediatric Neurology, Neurology, Child's Nervous System and Epilepsy and Behavior, indicating a greater interest in TSC's clinical neurology aspects, particularly epilepsy and neuro-behavioural/neuro-cognitive manifestations.