Palo Alto,  CA 
United States
  • 513

AI-powered genomic analysis and interpretation platform


Emedgene is an AI-based genomic analysis and interpretation platform applying technology to help geneticists manage their growing workload faster and with higher accuracy.

Emedgene's AI-powered genomics engine generates an accurate shortlist of potential causative variants along with their supporting evidence from the literature and databases, significantly reducing the time interpret a case. The automated interpretation machine learning algorithms have been trained on thousands of patient samples and incorporate dozens of features typically used by geneticists to interpret cases. The algorithms use a proprietary knowledge graph of variants, genes, diseases, phenotypes, and connections, which includes information extracted from literature with Natural Language Processing. Overall, the system automates time-intensive aspects of variant analysis and variant research and enhances lab analysis capabilities. 

In a study of 180 previously solved whole-exome sequencing cases (57 singles, 123 trios) submitted to Emedgene's platform, the algorithm successfully identified the causative variant (in a shortlist of 10 variants) in 96% of cases.  

 Press Releases

  • Applying its new Pathorolo algorithm, Emedgene will monitor patient cases and alert patients when cases can be solved with newly available clinical information

    PALO ALTO, CaliforniaFeb. 28, 2020 /PRNewswire/ -- Emedgene, an AI company working with the world's leading research institutes, is launching an initiative to help patients with undiagnosed rare diseases. The company will use its top-of-the-line genomics analysis platform and new Pathorolo algorithm to re-run patient cases and identify whether they can be solved with newly available information.

    Pathorolo is the first machine-learning algorithm developed to assess the likelihood of solving a genomic clinical case with currently available evidence. The model is based on Emedgene's proprietary automatic-interpretation algorithms, which include information continuously curated from the literature and databases. In a test cohort of 553 patients 93% of cases that were identified as solved, were indeed solved.

    Emedgene will continuously run its algorithms on the data and alert patients once cases are recognized as solvable. The recommended follow-up will be through a certified medical geneticist, who will be provided with the new evidence for the case. Emedgene will collaborate with patient advocacy groups, research institutes, and non-profit organizations to make the program available for their patients.

    One of the organizations that worked closely with Emedgene over the past year is the Rare Genomics Institute. Debora Varon, a Senior Genetics Analyst, attests to the benefits of automatically reanalyzing patients cases with novel scientific approaches: "I'm extremely excited about the hope this new capability brings to our community of undiagnosed patients."

    "Fewer than 50% of rare disease cases are solved on initial analysis," says Einat Metzer, CEO of Emedgene. "But genomics knowledge is growing at a fast pace, and at re-run, an estimated 10% of cases can be solved with new information. We'd like to close the gap between scientific knowledge and patient care. In this era of extremely rapid discovery, ironically, computational AI approaches can demonstrate human compassion."