A debate still exists concerning the role of occupational and environmental factors in the pathogenesis of systemic scleroderma (SSc).Our aim was to explore associations between SSc and environmental factors utilizing… Click to show full abstract
A debate still exists concerning the role of occupational and environmental factors in the pathogenesis of systemic scleroderma (SSc).Our aim was to explore associations between SSc and environmental factors utilizing an automatic semantic interpretation of PubMed results.The literature search string: (“systemic sclerosis” OR “scleroderma”) AND (“occupational exposure” OR “environmental” OR “risk factor”) was used to retrieve abstracts from the entire PubMed database, using Semantic MEDLINE 2, on 6/14/2020. This application represents a network of semantic predications (triples of the form subject-predicate (or relation) -object, e.g. Occupational Exposure causes Systemic Scleroderma) on a knowledge graph. Subject and object arguments of each predication are concepts from the Unified Medical Language System (UMLS) Metathesaurus and the relation is taken from the UMLS Semantic Network. The system allows for choosing the central topic (“Systemic Scleroderma”), the length of the network (3 nodes), and automatic summarization, eliminating the less informative predications.The search string retrieved 864 citations and identified 6,397 predications by using 34 types of relations. Initially, we focused our attention on the ‘CAUSES’ type of relation (Figure 1), displaying a network with 59 nodes and 57 edges.The central concepts of this network, identified as having causal relationship with SSc are autoimmune diseases/autoimmunity, chemicals such as bleomycin, occupational and environmental exposure, especially silica, vinyl chloride and trichloroethylene, genes, including HLA and non-HLA genes, genetic polymorphisms, transcription factors (TFs) such as Fli1 and KLF5, and fibrosis. Eosinophilia-myalgia syndrome, toxic oil syndrome and infection were all causally linked to autoimmune diseases. Minerals were associated with occupational exposure and with autoimmune diseases. Concepts causally linked to fibrosis were rare diseases, HLA genes, other non-HLA genes, such as STAT4, IR4, IR5, TLR4, TLR7 and Rho-associated Kinase, and vinyl chloride monomer. Pathogenic factors associated with SSc were endothelial dysfunction and extracellular matrix proteins. Many of the papers in the network also suggested that hormonal factors are involved.Inspection on the knowledge graphs reveals concepts central to research on the etiopathogenesis of SSc. The relations in which these concepts participate, provide more specific information. The Semantic MEDLINE graph supports the kind of patterns that underpin literature-based discovery.Although the pathogenesis of SSc remains elusive, it is accepted that initial vascular damage driven by autoimmunity and environmental factors causes abnormalities in the vasculature resulting in the activation of fibroblasts in various organs. Silica and solvents such as trichloroethylene seem to be the most consistently suspected environmental agents in SSc.[1]Rindflesch TC,et al. Semantic MEDLINE: An advanced information management application for biomedicine. Information Services & Use 2011;31:15-21.Figure 1.Semantic Network of Casual Relationships of Systemic Scleroderma.None declared
               
Click one of the above tabs to view related content.