Paradigm4 Gives Researchers Rapid Access and Scalable Analysis for Complex Scientific Data on AWS

For researchers to analyze population-scale, longitudinal scientific data quickly, they need to develop a scalable, cost-effective, and high-performance analytics platform that can handle genomic and phenotypic data.

Paradigm4, a science-driven data analytics platform, produces an agile, end-to-end solution that enables scientists to quickly analyze their data. Paradigm4 has joined forces with Alnylam Pharmaceuticals, a drug company dialing in on genetics capability, to analyze large-scale, high-content longitudinal scientific data contributed by 500,000 participants in a massive genomics sequencing project called the UK Biobank. With an ever-evolving and expanding data set, Alnylam needs to synthesize information interactively via an analytics platform that can scale and adapt affordably.

Paradigm4’s SciDB, built with multidimensional and scientific data in mind, meets Alnylam’s computing needs by leveraging Amazon Web Services (AWS) products. SciDB runs on Amazon Elastic Compute Cloud (Amazon EC2) instances—reducing costs using Amazon EC2 Spot Instances—and stores data accessibly using Amazon Simple Storage Service (Amazon S3). On AWS, Paradigm4 is able to run one billion linear regressions in less than an hour and one billion logistic regressions in a day or less. “That’s really the benefit,” says Alex Poliakov, Paradigm4’s vice president of customer solutions. “The scientists can get their results quickly. And because we use Amazon EC2 Spot Instances, the total cost that they pay is a lot less.”