Google partnered with the University of California Santa Cruz Genomics Institute to create an artificial intelligence-powered method that accelerates genome sequencing data analysis for faster disease detection.
The method, called the PEPPER-Margin-DeepVariant, analyzes data for Oxford Nanopore sequencers, one of the fastest commercial sequencing technologies available, according to a Jan. 13 company blog post.
On Jan. 12, the New England Journal of Medicine published a study led by the Stanford (Calif.) University School of Medicine demonstrating the method's use in a newborn intensive care unit. The method was used to analyze genomic data for 12 patients, and the researchers gave five diagnoses based on its analysis, allowing for more timely and disease-specific treatment.
In the fastest cases, a suspected disease-causing genomic variant was identified less than eight hours after sequencing began, compared with the previous record of 13.5 hours.
Google's post said new methods for genome sequencing data analysis can lead to breakthroughs. To further explore how its machine-learning technology can improve the field, the tech giant has partnered with Pacific Biosciences, a genomic sequence platform developer, to create more genome sequencing initiatives.