Image: Google has announced the open source release of DeepVariant, a deep learning technology to reconstruct the true genome sequence from HTS data with significantly greater accuracy than previous classical methods (Photo courtesy of Google).
Google LLC (Mountain View, CA, USA) has announced the open source release of DeepVariant, a deep learning technology to reconstruct the true genome sequence from high-throughput sequencing (HTS) data with significantly greater accuracy than previous classical methods. DeepVariant, which was developed after more than two years of research by the Google Brain team, in collaboration with Verily Life Sciences, transforms the task of variant calling, as this reconstruction problem is known in genomics, into an image classification problem well-suited to Google's existing technology and expertise. Within a year of its development, the deep learning model had won the 2016 PrecisionFDA Truth Challenge award for Highest SNP Performance, outperforming state-of-the-art methods. Since then, the Google Brain team has reduced the error rate by over 50%.
Google has released DeepVariant as open source software to encourage collaboration and accelerate its use in solving real world problems. It has partnered with Google Cloud Platform (GCP) to deploy DeepVariant workflows on GCP in two configurations optimized for low-cost and fast turnarounds using scalable GCP technologies, such as the Pipelines API. This paired set of releases provides a smooth ramp for users to explore and evaluate the capabilities of DeepVariant in their current compute environment, while providing a scalable, cloud-based solution to satisfy the needs of even the largest genomics datasets.
Google hopes that DeepVariant will leverage its computing infrastructure and machine learning expertise to better understand the genome as well as provide deep learning-based genomics tools to the community. The release of DeepVariant as open source software is part of its broader goal to apply Google technologies to healthcare and other scientific applications, and to make the results of these efforts broadly accessible.