Quantitative mapping of genetic similarity in human heritable diseases by shared mutations. Issue 2 (21st November 2017)

• Record Type: Journal Article
• Authors: Zhao, Huiying; Yang, Yuedong; Lu, Yutong; Mort, Matthew; Cooper, David N.; Zuo, Zhiyi; Zhou, Yaoqi
• Is Part Of: Human mutation. Volume 39:Issue 2(2018)
• Page Start: 292
• Page End: 301
• Publication Date: 2017-11-21
• Subjects: candidate gene -- disease similarity -- disease‐causing mutations -- disease‐susceptible individuals; Human chromosome abnormalities -- Periodicals; Mutation (Biology) -- Periodicals; 616.04205
• DOI: 10.1002/humu.23358
• Languages: English
• ISSNs: 1059-7794

Abstract

Abstract: Many genetic diseases exhibit considerable epidemiological comorbidity and common symptoms, which provokes debate about the extent of their etiological overlap. The rapid growth in the number of known disease‐causing mutations in the Human Gene Mutation Database (HGMD) has allowed us to characterize genetic similarities between diseases by ascertaining the extent to which identical genetic mutations are shared between diseases. Using this approach, we show that 41.6% of disease pairs in all possible pairs (42, 083) exhibit a significant sharing of mutations ( P value < 0.05). These mutation‐related disease pairs are in agreement with heritability‐based disease–disease relations in 48 neurological and psychiatric disease pairs (Spearman's correlation coefficient = 0.50; P value = 3.4 × 10 −5 ), and share over‐expressed genes significantly more often than unrelated disease pairs (1.5–1.8‐fold higher; P value ≤ 1.6 × 10 −4 ). The usefulness of mutation‐related disease pairs was further demonstrated for predicting novel mutations and identifying individuals susceptible to Crohn disease. Moreover, the mutation‐based disease network concurs closely with that based on phenotypes. Abstract : Disease‐disease relation can be captured more accurately using shared mutations than using shared genes as genes often have multiple functions and different mutations may have different consequences. This hypothesis is confirmed by several tests based on heritability, over‐expressed genes in cancer, identification of novel disease‐causing genes, and consistency with phenotypic classification.