Dr Wall and Dr Tonellato to give a keynote at Amazon Reinvent Conference
Dennis P. Wall (HMS) and Peter J. Tonellato (HMS, UWM, ZSPH) are invited speakers in the Amazon sponsored event “re:Invent” (November 12th-15th, Las Vegas) presenting a keynote lecture titled “The promise and problem of translational genetics and a step to the clouded solution of scalable clinical whole genome sequencing.”
They are participating in the “Big Data + HPC” track which is designed to provide an overview of the promise that ‘data has the opportunity to transform teams and build businesses, to drive innovation and help teams iterate more quickly.
In the Big Data + HPC track, you’ll hear from experts in data analytics, databases, storage, and high performance computation (HPC) and discover how to put data to work in your own organization. These sessions will help you learn how to run Big Data or HPC workloads on AWS, dive deep into Hadoop and Elastic MapReduce, or pro-tune your GPU code.’
Details of the lecture:
The problem and promise of translational genetics and a step to the clouded solution of scalable clinical whole genome sequencing.
Harvard Medical School in collaboration with Beth Israel Deaconess Medical Center have created a clinical-time, whole genome analysis workflow and management system. COSMOS is a parallelized workflow environment robustly deployed on AWS. Professors Wall and Tonellato from Harvard Medical School discuss the emerging area of clinical whole genome sequencing analysis, tools and tales of the use of AWS to achieve a robust 'clinical time' processing and the barriers and resolution of producing clinical-grade whole genome results on the cloud. In the talk, we will describe the steps to and methods behind COSMOS, a parallelized pipeline for optimum processing of whole genome sequences in a scalable manner on the cloud. We will detail how the solution can take advantage of various Amazon Web Services to help guarantee stability and failure management, and to manage cost. We will show several use cases that demonstrate benchmark our AWS solution against local computing solutions and that demonstrate the time and capacity gains conferred by use of AWS. Our focus will be on massively parallel genomics computing using EC2 services, and advanced harvesting of variable instances including spot instances for scalable growth. </blockqoute>