Alzheimer's disease (AD) is the leading cause of dementia and the most common neurodegenerative disorder, affecting over 5.5 million people in United States and 47 million people worldwide. Neuropsychiatric symptoms (NPS) are common in dementia and lead to great functional impairment, excess morbidity and mortality, and high social-economic costs in patients with AD. Currently, the genetic links between NPS and AD remains poorly characterized and there exist no effective treatments for AD and associated NPS. In this project titled ?Construct large-scale phenomes of diseases and drugs and develop combined phenome- and genome-driven systems approaches to understand genetic links between Alzheimer?s disease (AD) and Neuropsychiatric symptoms (NPS)?, we will develop natural language processing (NLP), information extraction, knowledge representation and data integration techniques to construct a large-scale MindDiseasePhenome from 28 million biomedical articles. MindDiseasePhenome contains deep phenotypic relationships for tens of thousands of diseases, including AD and many neuropsychiatric diseases. MindDiseasePhenome is standardized using biomedical ontologies, fully annotated with biomedical literature, and extensively integrated with existing disease genetics and genomics data. We will develop NLP, information extraction, knowledge representation and data integration techniques to construct a large-scale MindDrugPhenome from 28 million biomedical articles, > 36,000 FDA drug labels, and 13 million health records. MindDrugPhenome contains side effect (SE)/phenotype information of over 8,580 drugs and 22,499 SEs (including 1,898 neurologic and 740 psychiatric SE terms). MindDrugPhenome is standardized, annotated and fully integrated with existing genetics data and biomedical ontologies. We will innovatively leverage MindDiseasePhenome and MindDrugPhenome and develop data-driven systems approach to understand genetic links between AD and NPS. We will develop interactive web applications to make all the data publicly available to the broad scientific community through AMP-AD Knowledge Portal. The unique and powerful strength of our project is our ability to seamlessly combine automatic construction of large-scale phenomes of diseases and drugs with novel data-driven systems approaches to understand genetic links between AD and NPS. Our research integrates multiple advanced computational technologies, including NLP, information extraction, knowledge representation and reasoning, data integration, network biology and data-driven genetics predictions. The unique MindDiseasePhenome and MindDrugPhenome will set the foundation to enable other computational approaches to understand the genetic links between AD and NPS. Our predicted novel genetic links between AD and NPS will enable biomedical researchers to conduct hypothesis- driven functional studies in experimental models of AD. Our project will likely lead to discovery of novel genetic links between AD and NPS with translational implications for the development of therapeutic interventions.
! Alzheimer's disease (AD) is the leading cause of dementia and the most common neurodegenerative disorder, affecting over 5.5 million people in United States and 47 million people worldwide. Neuropsychiatric symptoms (NPS) are common in dementia and lead to great functional impairment, excess morbidity and mortality, and high social-economic costs in patients with AD. Currently, the genetic links between NPS and AD remains poorly characterized and there exist no effective treatments for AD and associated NPS. In this project titled ?Construct large-scale phenomes of diseases and drugs and develop combined phenome- and genome-driven systems approaches to understand genetic links between Alzheimer?s disease (AD) and Neuropsychiatric symptoms (NPS)?, we will develop advanced computational techniques to construct large- scale disease phenome (MindDiseasePhenome) and drug phenome (MindDrugPhenome) from 28 million biomedical articles, >36,000 FDA drug labels, and 13 million health records. We will leverage MindDiseasePhenome and MindDrugPheome and develop data-driven systems approach to understand genetic links between AD and NPS. The unique MindDiseasePhenomeKB and MindDrugPhenomeKB will set the foundation to enable other computational approaches to understand the genetic links between AD and NPS. Our predicted novel genetic links between AD and NPS will enable biomedical researchers to conduct hypothesis- driven functional studies in experimental models of AD.
| Zheng, Chunlei; Xu, Rong (2018) Large-scale mining disease comorbidity relationships from post-market drug adverse events surveillance data. BMC Bioinformatics 19:500 |
