Accelerating Cancer Diagnostics Pipelines Using Technologies of OpenPOWER

Abstract to be presented at Summit 2016

As DNA analysis makes its move from the research domain to actual clinical diagnostics applications, the urgency of reducing the time needed to perform such an analysis becomes ever more critical. There is a need to improve both the latency of the analysis, thereby reducing the overall time to patient diagnosis, as well as the throughput, thereby enabling to diagnose more patients using the existing computational infrastructure.

In this talk, we will present the collaborative effort by a number of OpenPOWER members, coordinated by the PersMed workgroup, to bring together various technologies from IBM, Xilinx, Nvidia, Nallatech, Bluebee and others, to accelerate the well-known Broad Best Practices pipeline (BWA/GATK), widely used for whole genome sequencing based diagnostics. The talk discusses software as well as hardware improvements to the pipeline, increasing the speed of various components by a factor ranging from 2x up to 50x. The talk also shows how these improvements can help reduce the time to diagnosis as well as help diagnose more patients in a clinical environment.

 

Speaker Bio

Zaid Al-Ars is cofounder of Bluebee, a company specialized in high-performance genomics solutions,  where he leads the research and development activities of the Bluebee offering. He is also an assistant professor at the Computer Engineering Lab of the Delft University of Technology, the Netherlands, where he leads the research and education activities of the multi/many-core research theme of the lab. Dr. Al-Ars spent a number of years in the Product Engineering Group of Infineon Technologies and Siemens Semiconductors in Munich, Germany, where he was responsible for constructing new test methodologies to reduce the overall cost of production test flows.