DeepVariant, Google's open source medical AI tool, is deployed in Google Cloud Google
DeepVariant, a tool built by Google Brain and Verily Life Sciences to improve the accuracy of genome sequencing, is now available through the Google Cloud Platform (GCP), laying the groundwork for widespread adoption of DeepVariant.
According to Healthcare IT News, the development team deployed the open source version of DeepVariant on the Google Cloud Platform with an extensible Pipelines API to optimize costs and achieve fast turnaround, allowing users to explore and evaluate DeepVariant's capabilities in today's computing environments , While providing a cloud solution that meets the needs of a huge data set of genomes.
DeepVariant addresses one of the biggest challenges in precision medicine today through the accuracy of high-throughput sequencing, developers DePristo and Poplin said that only 100 genes in the 3 billion base pair genome have a base error rate of 0.1 ~ 10%, all the paired bases are organized into 23 pairs of chromosomes.
To build DeepVariant, the Google Brain team first started with tens of millions of examples of coding high-throughput sequencing data starting with the reference genome organized by the Genome in the Bottle Consortium (GIAB) The next step was to train the Google TensorFlow classification model to distinguish between true genomes And experimental data.
In 2016, DeepVariant took the No. 1 spot in the PrecisionFDA Truth Challenge contest sponsored by the U.S. Food and Drug Administration (FDA) and has reduced the DeepVariant error rate by more than 50% since then.
Microsoft and the University of Pittsburgh Medical Center Collaborate to Develop Artificial Intelligence (AI) and Launve New Products That Include HealthVault Insights, Microsoft Genomics, Chat Robot and Project InnerEye to Establish an Artificial Intelligence Laboratory for Healthcare in the UK in September 2017 .
Since its acquisition of Phytel and Explorys in 2015, IBM Watson has focused on healthcare AI and added Watson for Genomic and Watson for Oncology to the supercomputer portfolio.
Google's goal is to apply technology to Medicare and other scientific applications that ultimately make tools and data available to more people DeepVariant leverages Google's computing power and machine learning to understand the genome and provide deeper insight into the community Learning-based genomics tools.