This technology is a tool for generating diagnostic genomic predictions based on electronic health record (EHR) data.
Unmet Need: Automated analysis of raw, unstructured EHR data for genetic diagnosis of rare diseases
Electronic health records provide a rich source of information on patient symptoms and can be useful for phenotype-driven diagnosis of genetic disease. However, existing phenotype-based gene prioritization computational tools are unable to process and interpret raw EHR data.
The Technology: Automated natural language processing pipeline for EHR data
This technology is a high-throughput EHR phenotype extraction and analysis pipeline. A natural language processing algorithm is used to convert EHR data into a format compatible with Phenolyzer, an existing software tool that can generate a prioritized list of potentially implicated genes based on phenotype data.
This software can enable comprehensive utilization of the wealth of data available within EHRs and facilitate the implementation of genomic medicine.
This technology has been successfully used to identify causative genes in two case studies, as well as a large scale pilot study.
Applications:
- Identification of disease-causing genes in patients
- Identification of genetic risk factors in patient populations
- Research tool for analyzing data in clinical studies
Advantages:
- Relies on readily available data
- Does not require restructuring of EHR data prior to analysis
- Increases efficiency of genetic testing by suggesting specific gene targets
Lead Inventor:
Kai Wang, Ph.D.
Patent Information:
Patent Status
Related Publications:
Son JH, Xie G, Yuan C, Ena L, Goldstein A, Huang L, Wang L, Shen F, iu H, Mehl K, Groopman EE, Marasa M, Kiryluk K, Haravi Ag, Chung WK, Hripcsak G, Friedman C, Weng C, Wang K. “Deep phenotyping on electronic health records facilitates genetic diagnosis by clinical exomes” Am J Hum Genet. 2018 July 5; 103(1): 58–73.30171-X)
Yang H, Robinson PN, Wang K. “Phenolyzer: phenotype-based prioritization of candidate genes for human diseases” Nat Methods. 2015 Sep; 12(9): pp. 841-843.
Tech Ventures Reference: