Efstratois Efstathiadis, Ph.D.

Technical Director, New York University Langone Medical Center

HPC Challenges in Next Generation Sequencing

High throughput Next Generation Sequencing (NGS) instruments are capable of producing millions of short DNA reads per single run at a fraction of the cost and time compared to pre-NGS technologies. The dropping cost of full genome sequencing is enabling scientists to sequence genomes at a massive scale, initiating many new projects, leading to better understanding of human variation and investigating personalized medicine. New industries are emerging as sequencing labs, service providers and drug makers are deploying NGS instruments. The management of enormous amounts of digitized sequence data, mining, analysis and knowledge extraction creates a demand for scalable, extensible and adaptable HPC resources, expertise and funding. Addressing such challenges is necessary in advancing personalized medicine, which enables the integration of patient's unique genetic profile in making informed clinical decisions, designing drugs and developing preventive healthcare.

The newly established HPC facility at NYULMC collaborates closely with the Genome Technology Center and the Sequencing Informatics Group to address the compute and data storage NGS needs. The facility has deployed locally an HPC compute cluster and substantial centralized data storage capacity enabling multiple data analysis applications and visualization tools, operating on different servers, to access centrally stored data, reducing data transfers over the local network and duplication of large data sets in multiple locations. Consultation services are provided to evaluate and benchmark external supercomputing resources and compute cloud options.

Biography

As the Technical Director of the High Performance Computing Facility at NYULMC I provide access to centralized computing resources and services to our research community. I review and recommend strategies for high throughput data analysis and scientific computations, develop algorithms and benchmarks and provide consultations on best practices and current trends in High Performance Computing. Prior to NYULMC I was a technology architect responsible for massively parallel computing architectures and worked on deploying scientific simulations on Grid and cloud computing infrastructures.

Workshop Program
updated: 2011-10-19