ANN ARBOR – Genomic Data Search company Genomenon has been awarded a Phase 2 Small Business Innovation Research grant by the National Institutes of Health. The amount was not disclosed.
The grant will be used to develop and apply machine learning and artificial intelligence algorithms to variant interpretation, the single biggest hurdle in scaling the adoption of genomic sequencing in both clinical and precision medicine applications. Genomenon will be applying advanced ML and AI algorithms to its Mastermind genomic search database to provide suggested interpretations of genomic variants.
“The NIH has been a key partner in helping us tackle the single biggest barrier at the center of genomic medicine – automating variant interpretation based on the scientific evidence from the medical literature” said Mike Klein, CEO of Genomenon. “We are excited about the results that we are able to produce based on the work funded by this grant.”
Genomenon has taken an innovative new approach when it comes to applying AI to genomic data sets. Many companies have applied natural language processing and ML to genomic literature with little success. Genomenon has avoided the pitfalls of NLP, as there is nothing “natural” about the dozens of different ways that authors may describe genetic variations in the scientific literature.
Instead, the company has developed and patented its proprietary Genomic Language Processing technique to find every disease, gene and variant in over 6 million full text genomic articles. Rather than applying ML and AI in a wholesale fashion to its search engine database, Genomenon uses specific GLP algorithms to refine the quality and priority of each individual search result. This leads to results that are both highly sensitive and highly specific in finding clinically relevant citations tied to a patient’s DNA.
The grant from NIH will help Genomenon further automate the literature curation process by applying GLP across American College of Medical Genetics and Genomics / Association of Molecular Pathologists (ACMG/AMP) guidelines to determine variant pathogenicity. Unlike some black-box solutions applying AI, Genomenon produces transparent results, displaying both the GLP evidence underpinning its conclusions and the citations and cited sentences that go into variant scoring.
Research reported in this publication was supported by the National Human Genome Research Institute of the National Institutes of Health under Award Number R44HG009474. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.