Cornell University MacCHESS Facility

Researcher Brings Open-Source Software to the Mac

Like most areas of genomics research, protein crystallography requires extreme computational speed. A protein crystallography researcher at Cornell University, Dr. Michael Love has not only found ample power in his Power Mac mini-cluster, he is also leading the way in bringing open-source research tools to the Mac platform. Through the GNU-Darwin project he founded, Dr. Love is providing free software to the many researchers who prefer Mac.

Protein crystallography seeks to discover how proteins carry out the mechanical and chemical workings of a cell. This field studies the structure of protein molecules in atomic detail, a process that involves collecting x-ray diffraction data from protein crystals. To maximize the diffraction, much of the data is collected using high-intensity x-ray beams at large synchrotron facilities such as CHESS in Ithaca, New York, which hosts MacCHESS, a facility for macromolecular diffraction studies. A protein crystallographic data set can consist of thousands of measured intensities that are a mathematical representation of the atomic coordinates of the protein molecule under study.

“With the availability of Mac OS X, I can do everything on my preferred computer — the Mac.”

— Dr. Michael Love, Postdoctoral Associate,
Cornell University Department of Molecular Biology and Genetics

The More Speed, The Better

Progress in protein crystallography has traditionally been limited by available computational power. Billions of calculations may be required to translate the crystallographic data into a protein structure. More computational speed is always better, which is why many protein researchers have turned to techniques such as parallel processing.

In the past, Dr. Love conducted much of his protein crystallography research on UNIX workstations because of their parallel processing abilities. At the same time, he has always preferred the elegance and ease of use of the Macintosh. With the availability of Mac OS X, Dr. Love can harness the power of UNIX and the intuitiveness of the Macintosh on one platform. “I’m a longtime Mac user and also somewhat of a UNIX aficionado,” says Dr. Love. “With the availability of Mac OS X, I can do everything on my preferred computer — the Mac.”

PyMOL runs under Max OS X for studying protein DNA and protein drug interactions.

Clustering for High-End Compute Power

In his lab, Dr. Love runs two dual-processor Power Mac computers as a parallel processing mini-cluster. At home, he has an Apple iBook and another Power Mac. The cluster gives him all the power and speed he needs for protein crystallography, while costing far less than high-end UNIX workstations.

When he first set up his lab, Dr. Love needed to build a specialized computational infrastructure in preparation for his own research complete with shells, common UNIX applications, scripting languages, and compilers for software development. To fill the temporary gap for Mac-based tools until researchers port more scientific software to Mac OS X, Dr. Love began creating GNU-Darwin, a free standalone UNIX operating system for the Mac based on the Darwin UNIX software core with an X11-based user interface. By running both Mac OS X and GNU-Darwin on his systems, Dr. Love has both productivity software and all of the existing UNIX research tools he needs on the Mac.

Dr. Love uses his Power Mac mini-cluster for solving protein structures, crystallographic refinement, molecular modeling and building, and publication-quality molecular graphics production. Key applications on the Mac such as CNS, PyMOL, and CCP4 provide a comprehensive tool set for his research.

1 2 Next Page >