Zeblok Computational Announces Partnership with Laufer Center – COVID-19 Drug Discovery
Mon Jul 20 2020 11:00:00 GMT+0000 (Coordinated Universal Time)
STONY BROOK, NY, JULY 20, 2020 – Zeblok Computational is contributing to the efforts of two groups of researchers at The Laufer Center for Physical and Quantitative Biology, for COVID-19 drug discovery by providing Zeblok’s secure, scalable Supercomputing-as-a-Service platform, employing 128 GPUs from the CEWIT SMART GPU Cluster.
The Laufer Center simulations focus on 3 key proteins of the SARS-CoV-2 virus to pinpoint more precisely which drug candidates have the best binding affinity to various protein targets:
1. The RNA-dependent RNA polymerase (which copies the virus inside the cell, thought to be the target
of Gilead's Remdesivir, now in clinical trials)
2. The Protease (which chops up the virus proteins into the right pieces)
3. The spike protein (which is how the virus unlocks the door to enter the cell).
The Laufer Center is running their computationally-intense simulations to understand how proteins fold and how proteins bind to proteins – actions, mechanisms and binding affinities – to enable design of drugs for high-affinity binding to proteins. They apply state-of-the-art modeling, using physics-based methods to examine motions, forces and free energies within biological mechanisms. In a previous limited benchmarking program, the Laufer Center concluded that they achieved a 6x improvement in performance on Zeblok’s platform compared to NSF’s Blue Waters Cray GPU-based system. The structures of most COVID virus proteins are not yet known. Dr. Ken Dill’s group is computing many of their structures using MELD x MD, while Dr. Dima Kozakov’s group is computing structures of human-virus and virus-virus protein pairs.
“Companies offering in silico drug discovery services, such as computer-aided drug design (CADD), molecular modeling and quantitative structure-activity relationship (QSAR), have now become an important part of the pharmaceutical industry. Drug developers, especially those focused on the development of large molecules, are likely to continue relying on outsourcing for a significant part of their respective drug discovery and development operations. Zeblok Computational platform levels the playing field for researchers that need accelerated infrastructure for R&D. We uniquely combine computing platforms, orchestration, multi-cloud and in silico software developers to deliver quality services for companies engaged in drug discovery” said Zeblok Computational CEO Mouli Narayanan.
Zeblok Computational is incubated at and supported by CEWIT, using its SMART cluster of 180 NVIDIA RTX 6000 GPUs. “Our dual-use homogenous SMART Cluster is one of the fastest GPU clusters in academia and supports dual-use of both deep learning and visualization in a variety of applications” said Dr. Arie Kaufman, the Chief Scientist and the SMART Cluster Director. Mr. Narayanan further said “We want to help data center operators at the grass roots level to offer AI and AI-Hyperconvergence capabilities to their customers, enabling them to participate in the AI ecosystem. Companies are confronted with so much data that they increasingly need AI enabled smart processes. Implementation of 5G is further driving the low latency requirement for AI-Inferencing at Edge data centers. Zeblok’s orchestration engine seamlessly integrates multiple disparate resources into a complete AI model development environment all the way to the Edge, supported by its smart sourcing business model for surplus GPU capacity.”
About Zeblok Computational
Zeblok Computational Inc. is an Artificial Intelligence Platform-as-a-Service, providing a single unified pre-configured environment, with familiar frameworks and languages as well as a growing library of carefully selected proven original AI algorithms, enabling data scientists to begin developing AI models in minutes, to scale effortlessly for model training and to deploy in a runtime environment, with AI expertise and high touch assistance in integrating AI solutions into enterprise business processes. Zeblok’s container-based engineering platform leverages familiar open source software, adding a sophisticated proprietary multi-class, multi-cloud orchestration engine that combines heterogeneous environments through abstraction and virtualization and includes built-in resource management, multi-tier workload distribution, both front-end and back-end services and scale-out resource availability in minutes. Zeblok’s secure, smart sourced, Edge-connected accelerated infrastructure comprises a multi-cloud of over 7,000 GPUs, enabled by partnerships with the CEWIT and others. For more information: computational.zeblok.com
About The Laufer Center
The Louis and Beatrice Laufer Center, established in 2008 to advance biology and medicine through discoveries in physics, mathematics and computational science, is a hub for Physical and Quantitative Biology research at Stony Brook University. Laufer Center researchers come from several Stony Brook departments and Cold Spring Harbor Laboratory. For more information:
laufer-covid.org and laufercenter.stonybrook.edu
New York’s Center of Excellence in Wireless and Information Technology (CEWIT) at Stony Brook University is a next-generation research and educational facility with a mission to conduct interdisciplinary applied research in emerging and critical technologies, address the skilled technology worker shortage and to foster new enterprise development in wireless and IT communications, infrastructure, networking, and software. When CEWIT upgraded its accelerated infrastructure to a smart cluster of over 180 NVIDIA RTX 6000 GPUs in 2019, they realized they needed to orchestrate the various resources to achieve seamless distribution of compute power throughout the university, which led to its partnership with Zeblok. For more information: cewit.org