Dynamics and Emerging Trends in Genodermatology: A Scientometric Analysis

Abstract

With the improved power and precision with genome sequencing, the body of relevant literature on genodermatology has grown rapidly in recent years. In this study, we used CiteSpace, a scientometric software, to delineate the structure and dynamics of genodermatology and to further understand the emerging trends in this field. CiteSpace integrates individual citations and forms a co-citation network that demonstrates the evolution of a scientific field over time. Each reference cited by multiple articles is represented by a node in the network. The size of a node indicates how many times the reference has been cited. The nature of the connectivity between the nodes represents a dual relationship between the cited references and their citing articles. Networks formed in this way reveal the focuses of scientific research.1 We searched Web of Science on April 28, 2021 to identify literature published between 2000 and 2021 using the topics “genodermatosis” OR “genodermatoses” OR “genetic skin diseases” OR “genetic skin disease” OR “genetic dermatosis” OR “genetic dermatoses” OR “inherited skin disease” OR “inherited skin diseases.” This search retrieved 1,418 records (1,201 original research articles and 217 review articles). The citation index-based expansion resulted in the identification of 13,838 records, including 10,398 original research articles and 3,440 review articles. After filtering out repeated records, the dataset for analysis was reduced to 10,857 original research articles and 3,481 review articles. The analysis of publication timeline on genodermatology showed that the literature network was divided into 39 distinct co-citation clusters, including 15 main diseases, 19 target genes, and five methods within the past two decades (Fig. 1). Among the 39 clusters, Cluster #0 (epidermolysis bullosa [EB]) and Cluster #5 (atopic dermatitis) are of particular interest. It was evident that Cluster #0 had a high enrichment of nodes with citation bursts, which attests to the fact that this was the most striking cluster since 2004. Cluster #5 had citation bursts since 2010, suggesting that it was the most recently formed cluster. Cluster #1 (Birt-Hogg-Dube syndrome) had continuing publications with citation bursts for almost 20 years; however, the concentration of nodes has reduced since 2014. Clusters #2–4, #6, #7, #18, #20, #22, #37, #45, #48, #53, #58, and #63 had no citations after 2013. In contrast, Clusters #8–10, #19, #21, #24, #28, #30–33, #36, #50, and #55 mostly developed after 2007 or within the past 10 years; however, none of these clusters had nodes with citation bursts in the past 5 years, suggesting that there are no potential emerging trends in these clusters.Figure 1: Timeline of co-citation clusters. The size of a node indicates how many citations the associated reference received. Each node is depicted as a series of citation tree-rings across the series of time slices. Citation rings in red indicate the time slices in which citation bursts, or abrupt increases in citations, are detected. Citation bursts provide a useful means to trace the development of research focus. CiteSpace divides the co-citation network into clusters of co-cited references, so that references are tightly connected within the same clusters, but loosely connected between different clusters. Modularity Q has a value of 0–1, with Q > 0.3 indicating that the network community structure is significant; the current Modularity Q = 0.9376. Silhouette S > 0.7 indicates that the clustering result has high confidence. The closer the Silhouette S is to 1, the more homogeneity the network has; the current Silhouette S of the network is 0.9751.For Cluster #0, the two most significant citation nodes were in 2008 and 2014 and their corresponding literature is “The classification of inherited EB: Report of the Third International Consensus Meeting on Diagnosis and Classification of EB” by Fine et al.2 (burst value = 88.81, number of citations = 205) and “Inherited EB: Updated recommendations on diagnosis and classification” by Fine et al.3 (burst value = 86.47, number of citations = 203). Since the EB classification system was published in 2000, Kindler EB (old name Kindler syndrome), caused by mutations in the FERMT1 gene (old name KIND1 gene), and a series of pathogenic gene loci, such as plakophilin 1 (PKP1) and desmoplakin (DSP) for suprabasal EB simplex, were added into its arsenal in 2008. One contemporaneous review focusing on the genetic classification progress in EB4 also received substantial attention (burst value = 12.81, number of citations = 28). Specifically, the period from 2000 to 2008, there was a citation burst caused by research articles investigating the therapeutic effect of the transplantation of genetically modified epidermal stem cells on EB (burst value = 30.76, number of citations = 68)5 and bone marrow (burst value = 29.24, number of citations = 66).6 The updated EB classification in 2014 includes new targeted genes such as transglutaminase 5 (TGM5), junction plakoglobin (JUP), integrin alpha3 (ITGA3), and clinical subtypes such as suprabasal EB simplex and junctional EB with respiratory and renal involvement. During the period from 2008 to 2014, a citation burst was caused by articles focusing on the treatment effects of pluripotent stem cells (burst value = 21.97, 30.32, and 31.45; number of citations = 48, 69, and 80)7–9 and a precise genome editing technique (burst value = 22.94, number of citations = 28).10 In the subsequent years, publications with a large number of citations focused on the treatment effects of transgenic stem cells (burst value = 41.69, number of citations = 81)11 and traditional autologous epidermal grafts (burst value = 29.57, number of citations = 62)12 on EB. Cluster #5 comprised the research outcomes of atopic dermatitis, a complex disease that is now considered as a polygenic skin disease. Atopic dermatitis is the only polygenic genetic disease included in this analysis, and studies of the genetic contributions to atopic dermatitis have attracted much attention in recent years. The citation bursts of this cluster were caused by articles investigating the roles of skin barrier dysfunction (burst value = 57.57, number of citations = 127),13 epithelial cell-derived cytokines (burst value = 17.53, number of citations = 39),14 and filaggrin mutations (burst value = 16.24, number of citations = 37)15 in atopic dermatitis. We further analyzed the modularity of the whole co-citation network to identify publications introducing profound structural variation, which may lead to emerging trends. Fig. 2 shows the change in the modularity of networks over time; the modularity dropped in 2020 and returned to the previous level in 2021, suggesting that groundbreaking articles were published in 2020. Considering both the modularity change and citation data, the review article entitled “Consensus reclassification of inherited epidermolysis bullosa and other disorders with skin fragility”16 in Cluster #0 may be the driving force. This consensus introduced the concept of genetic disorders with skin fragility, of which classical EB represents the prototype. The proposed classification scheme with new concepts substantially changed and expanded the disease spectrum, which has the potential to create emerging trends in the future. The long-term citation impact of this publication warrants further investigation.Figure 2: The modularity of the network of co-citation references over the past 20 years. There is a considerable decrease in 2020 and a return to the previous level in 2021. Modularity Q has a value of 0–1, with Q > 0.3 indicating that the network community structure is significant. The current Modularity Q of the network is 0.9376.In conclusion, we performed scientometric analysis to identify the research structure and dynamics of genodermatology in the past twenty years. Our scientometric analysis showed that the main focus of investigation is EB and associated disorders related to skin fragility. With the advent of next-generation sequencing with high throughput sequencing power and at a plummeting cost, mutations in known candidate genes and previously unrecognized genes will continue to be disclosed for Mendelian and polygenic skin diseases. Unraveling their molecular basis will enhance our understanding of the disease pathogenesis and will provide insights into therapy development, with translational implications to more common, acquired skin conditions. Source of funding This work was supported by China Science and Technology Journal Excellence Action Plan (C-043) and China Science and Technology Journal Excellence Action Plan Selecting and Cultivating High-level Journal Talents Sub-Project-Youth Talent Support Project (2020ZZ110402).