At the annual national meeting of the American Chemical Society in Philadelphia, the hopes, frustrations and successes of building an instant, temporary supercomputer were generated once again.  FlashMob I’s previous attempt in the spring of this year proved the concept of grassroots supercomputing.  This time, FlashMob had scientific success that potentially benefits the safety and security of the United States.

A group of computational chemists, including organizers from FlashMob I (www.flashmobcomputing.org) like Pat Miller of Lawrence Livermore National Laboratory’s Center for Applied Scientific Computing, joined together to build a supercomputer in only a few hours.  The cooperative system was then used to study a protein-protein interaction that is involved in the activation of anthrax infection.

Because of the expense and time involved in the permanent assembly of traditional supercomputers, which require professionally trained staff to construct and continually run, they are usually built by government laboratories, national research universities or corporations.  This project, as temporary as it was, however, was very inexpensive and took only a few hours to assemble, use and take apart.  With such instant supercomputing projects, smaller universities, community colleges and high schools can all get involved in building their own supercomputers.

The demonstration was sponsored by Semichem, a software company, that helped fund the joint project between the Division of Computers in Chemistry and the organizers of FlashMob I.  This attempt at instant supercomputing was the first time a FlashMob computer was assembled specifically to solve a scientific problem.

Jeffrey Evanseck, a chemistry professor at Duquesne University, suggested the anthrax infection problem to the supercomputing team.  The spores of anthrax have been used (or have been threatened to be used) as a biological weapon against the military and civilian population.  Protection against anthrax infection has become a national homeland defense priority in recent years. The goal of this FlashMob demonstration was was to use the instant supercomputer to investigate the interaction of human proteins involved in activating anthrax infections.

Michelle Francl, chemistry professor at Bryn Mawr College, Miller, and the rest of the team used two miles of cable and dozens of borrowed laptops to solve the previously unsolvable anthrax infection problem in just forty minutes.  Supercomputing proved to be not only a successful concept, but also a successful tool for solving scientific problems.