Technology developed to study the stars is being applied to more earth-bound subjects in recent agreements that place astronomyÂ’s tools in the hands of bioinformatics researchers.
In October, the Association of Universities for Research in Astronomy (AURA) entered an agreement with Celera Genomics for the use of AURAÂ’s Operational Pipeline Unified Systems (OPUS) software package.
OPUS was originally designed by the Space Telescope Science Institute (managed by AURA under contract with NASAÂ’s Goddard Space Flight Center) to process electronic data for the Hubble Space Telescope program. It orchestrates the steps of the pipeline that converts raw, noisy data to processed images. Celera intends to use it to process its pipeline of genomic and proteomics sequence data.
Meanwhile, NASAÂ’s Center for Computational Astrobiology and Fundamental Biology (NCCAFB), based at NASAÂ’s Ames Research Center, and StanfordÂ’s Center for Biomedical Computation (CBMC), have entered a partnership to develop new computational biology tools to study how cells function, evolve, and are affected by diseases on Earth and in space.
Stanford will immediately benefit from AmesÂ’ role as NASAÂ’s primary information technology center. The NCCAFB currently uses nearly 4,000 processors on SGI supercomputers at the NASA Advanced Supercomputing Division at Ames. The Stanford researchers will have access to this system under the partnership, although terms of access have not yet been finalized. The CBMC currently uses a SGI Origin 3800-class computer.
But Stanford won’t be the only partner to benefit from the collaboration. “[Stanford is] very strong in biological research and NASA is fairly new in the field of biology,” said Andrew Pohorille, director of the NCCAFB. “We have the benefit of working with somebody so strong.” NASA will also gain access to a steady stream of Stanford students to use as interns in its labor-intensive research projects.
Russ Altman, director of the CBMC, said initial projects would focus on 3D structural simulations of molecules and protein structure analysis and prediction, “because this is an area of strength at both NASA and at Stanford.”
Other projects will include the integration of diverse databases, the simulation of life processes, and the development of methods for moving between images and their corresponding physical models. The bulk of the research will focus on cell metabolism, using both healthy and diseased cells cultured from experiments on Earth and in space. In addition, the partnership will develop new information management tools to use on NCCAFBÂ’s massively parallel computers.
Altman said the Stanford/ NASA partnership arose from the realization that “the bioinformatics and computational biology world is not that big, and so we should be in communication with our colleagues who are right down the road.”
Celera’s tie-up with AURA, however, arose more from a chance circumstance. Steven Beckwith, director of the Space Telescope Science Institute, said AURA had no plans to license its OPUS technology, but “Celera approached us.”
Celera had been doing many of the processing steps performed by OPUS by hand. A newly hired programming engineer who had previously worked on astronomical data realized that OPUS, which is well-known in the astronomical world for data pipeline management, would easily make Celera’s raw data “fit for scientific consumption,” Beckwith said.
Because Celera’s data stream was around 10 times the size of the Hubble data stream — 50 GB vs. around 4 GB per day — both partners were pleasantly surprised to find that OPUS could handle the extra load.
Beckwith said that AURA would be willing to license the technology to other organizations that process large amounts of data.
— BT