M171. THE GENE-SHARING RELATIONSHIP OF SCHIZOPHRENIA WITH OTHER MENTAL OR SYSTEMIC DISORDERS: A DISEASE-SIMILARITY NETWORK ANALYSIS FOCUSED ON EGOCENTRIC NETWORK
Abstract Background Schizophrenia is an archetypal example that a psychiatric illness may not merely be a mental or a brain disorder but rather a systemic illness. It can be glimpsed from a wide range of biomarkers that span all the imaginable body systems, and from higher co-morbidity with other systemic illnesses. However, quantitative analysis of schizophrenia’s relationship with other diseases are not yet satisfactory. Genome-wide association studies have identified more than hundreds of genetic loci associated with schizophrenia. In turn, these loci are associated with a wide variety of other diseases. From this gene-disease relationship, a bipartite network can be built which, after appropriate projection, could help to map a complex disease-similarity network. In case of schizophrenia, it would reveal the position of schizophrenia among the broader categories of systemic illnesses. Methods DisGeNET is a discovery platform which contains one of the largest collections of gene-disease association data. The major source of the integrated data is the automatized curation from MEDLINE abstract. Therefore, it contains the timestamp of reported gene-disease association. Gene-disease-timestamp (year of publication) triplet was fed into a Neo4J graph database platform. From this, disease-disease relationships with shared gene count and Jaccard similarity score was extracted. The network structure of level 1.5 egocentric network centered upon schizophrenia was inspected. Louvain community detection algorithm was applied to expose underlying group structure among the 1st order alters. For comparison, similar ego-networks centered upon several major psychiatric illnesses were also inspected. Finally, the yearly variation of Jaccard score which reflected the accumulation of research data were monitored. Results The diseases which showed the highest Jaccard score (j) were bipolar disorder (j=0.203) and depressive disorder (j=0.190) as expected. Other diseases with meaningful similarity could be grouped into three communities: 1) psychiatric illness including bipolar/depressive disorder, 2) a variety of malignancies including neuroblastoma (j=0.083), stomach cancer (j=0.070) and pancreatic cancer (j=0.065) 3) other systemic illnesses including multiple sclerosis (j=0.088), metabolic syndrome (j=0.076), myocardial infarction (j=0.073), rheumatoid arthritis (j=0.070), lupus erythematosus (0.056). The gene-sharing relationship with systemic illnesses (malignancies and other) began to be revealed after 2005. Since then, more and more evidences were accumulated to solidify the schizophrenia’s link with systemic illnesses. Discussion Recently, a couple of large-scale epidemiological studies verified the significant correlation between prevalence of schizophrenia and cancer/autoimmune disorders. The present study results may augment these epidemiological data and thus strongly support the concept of schizophrenia as a systemic illness. Gene-sharing and its reflection in prevalence data would indicate deeper link at the level of pathogenesis with systemic illnesses. Recently, many authors contemplated the possible link between schizophrenia and cancer in terms of cell cycle regulation and control of apoptosis. Likewise, others suspected immunological disturbance as the fundamental mechanism of schizophrenia. In this vein, the need for extending the concept of mental disorders as a focused manifestation of systemic illness seems gaining impetus.
