Identifying Pleiotropic Variants Using ClinVar & National EHR Data
Mentor 1
Dr. Peter Tonellato
Location
Union Wisconsin Room
Start Date
29-4-2016 1:30 PM
End Date
29-4-2016 3:30 PM
Description
This work describes a novel method linking publically available genomic data with hospital discharge data to uncover pathological pleiotropic variants. When a variant arises in a pleiotropic gene, distinct diseases with different symptoms and clinical therapies can co-occur in a patient, obscuring the diseases’ shared genetic origin. Identifying pathological pleiotropic genetic variation has significant clinical implications for directing research towards targeted therapy and impacting clinical guidelines for screening and monitoring. In contrast to other techniques for pleiotropic discovery that require extensive study design and resources, we hypothesized that pathological pleiotropic genetic variants can be identified using existing genomic and phenotypic information in the National Center of Bioinformation Technology (NCBI) database ClinVar. We mined the ClinVar database for specific diseases, colorectal cancer and asthma, both of which have at least one known genetic component and are the source of significant morbidity and mortality in the US population. We found 1,120 variants linked to colorectal cancer risk; 53 variants remained after excluding those provided by a single submitter and those linked only to colorectal cancer. We identified specific variants linking colorectal cancer to other diseases with already established relationships such as Lynch Syndrome and breast cancer. We also uncovered a novel relationship between colorectal cancer and age-related macular degeneration (ARMD), linked by independent studies of the variant rs4986790 of the TLR4 gene. Similarly, we found 24 variants linked to asthma, and 1 variant, rs1042713 of the ADRB2 gene, was independently associated with both asthma and metabolic syndrome. These results support our hypothesis that ClinVar can yield valuable information on possible pathological pleiotropic variants. Further, we matched ICD-9 codes with diseases found to be associated in ClinVar. We then proceeded to test for epidemiological relationships between the associated ICD-9 codes by mining nationwide hospital discharge data. Preliminary results confirm an increase in relative risk for diseases genetically linked in ClinVar across a broad population and suggests a method for validating the pleiotropic relationships found in ClinVar. Further work is being implemented to systematically search ClinVar for pleiotropic variants and cross-validate them with ICD-9 codes in hospital discharge data.
Identifying Pleiotropic Variants Using ClinVar & National EHR Data
Union Wisconsin Room
This work describes a novel method linking publically available genomic data with hospital discharge data to uncover pathological pleiotropic variants. When a variant arises in a pleiotropic gene, distinct diseases with different symptoms and clinical therapies can co-occur in a patient, obscuring the diseases’ shared genetic origin. Identifying pathological pleiotropic genetic variation has significant clinical implications for directing research towards targeted therapy and impacting clinical guidelines for screening and monitoring. In contrast to other techniques for pleiotropic discovery that require extensive study design and resources, we hypothesized that pathological pleiotropic genetic variants can be identified using existing genomic and phenotypic information in the National Center of Bioinformation Technology (NCBI) database ClinVar. We mined the ClinVar database for specific diseases, colorectal cancer and asthma, both of which have at least one known genetic component and are the source of significant morbidity and mortality in the US population. We found 1,120 variants linked to colorectal cancer risk; 53 variants remained after excluding those provided by a single submitter and those linked only to colorectal cancer. We identified specific variants linking colorectal cancer to other diseases with already established relationships such as Lynch Syndrome and breast cancer. We also uncovered a novel relationship between colorectal cancer and age-related macular degeneration (ARMD), linked by independent studies of the variant rs4986790 of the TLR4 gene. Similarly, we found 24 variants linked to asthma, and 1 variant, rs1042713 of the ADRB2 gene, was independently associated with both asthma and metabolic syndrome. These results support our hypothesis that ClinVar can yield valuable information on possible pathological pleiotropic variants. Further, we matched ICD-9 codes with diseases found to be associated in ClinVar. We then proceeded to test for epidemiological relationships between the associated ICD-9 codes by mining nationwide hospital discharge data. Preliminary results confirm an increase in relative risk for diseases genetically linked in ClinVar across a broad population and suggests a method for validating the pleiotropic relationships found in ClinVar. Further work is being implemented to systematically search ClinVar for pleiotropic variants and cross-validate them with ICD-9 codes in hospital discharge data.
