The new Onco-SciDB (oncology scientific database) features a graphical user interface designed for exploring data from The Cancer Genome Atlas (TCGA) and other relevant public.
The Onco application runs on top of the Paradigm4’s SciDB database management system devised for analysing multi-dimensional data in the cloud. The management system was built by database pioneer Michael Stonebraker in order to use the cloud for massively parallel processing and offering an elastic supply of computing resources.
A cloud-based database system gives research departments cost control and the capacity to ramp up production when needed, according to Paradigm4 CEO Marilyn Matz. “The result is that research teams spend less time curating and accessing data and more time on interactive exploration,” she said.
Currently, the bioinformatics industry lacks the requisite analytical tools and user interfaces to deal with the growing mass of molecular, image, functional, and clinical data, according to Matz. By simplifying the day-to-day challenge of working with multiple lines of evidence, Paradigm4 claims that SciDB supports clinical guidance for programmes like precision anti-cancer chemotherapy drug treatment. By making massively parallel processing possible in the cloud, it claims, it can provide sufficient affordable computing power for budget-constrained research institutes to trawl through petabytes of information and create hypotheses over the various sources of molecular, clinical and image data.
Database management system SciDB serves as the foundation for the 1000 Human Genomes Project and is used by bio-tech companies such as Novartis, Complete Genomics, Agios and Lincoln Labs. A custom version of Onco-SciDB has been beta tested at cancer research institute Foundation Medicine.
Industry veteran Stonebraker, the original creator of the Ingres and Postgres systems in 1985 that formed the basis of IBM’s Informix and EMC’s Greenplum, won the Association for Computing Machinery’s Turing Award and $1million from Google for his pioneering of database design.