Forays Into "Citizen Science"
Some Research Is Getting Done Outside The Lab. Do You Want To Help?
By Amy Pletcher
When most of us think of research, we imagine a lone, highly-trained scientist huddled over his work in a laboratory late into the night. When we took science class in school, even group projects usually didn't exceed a half-dozen people. Today however, scientists are teaming up with literally millions of people to make progress on their projects. Thanks to David Gedye and others, you can team up with scientists on the cutting edge of fields ranging from astronomy and climate prediction to protein structure.
The first distributed computing system was founded in 1995 when David Gedye looked around his laboratory at the University of California in Berkeley, Calif. and realized that he was going to need help to get through all his data before it became outdated, and SETI@home was born.
SETI@home (Search for ExtraTerrestrial Intelligence) is a program that makes use of untapped computer resources from upwards of five million computers around the world to analyze transmissions for signs of extraterrestrial life. Gedye and his team designed a distributed processing system and software that would allow them to access the processing power of dormant computers from anyone that downloaded their software. These individual computers break up deep space radio telescope data into work units that take about an hour of processing time and then send information about any abnormalities back to the SETI mainframe in Berkeley. Currently, over 5 million users from over a hundred countries have signed up for SETI@home and have contributed over 19 billion hours of computer processing time.
In addition to information about the data being analyzed, the SETI@home website comes with all sorts of other places to explore. There are lists of users of the day (selected from top contributors), as well as lists of the top analyzers in a variety of categories. As of this writing, Mardonios Goudoulias in Cyprus is the top primary school team, with about 514,974 credits. There are also categories for secondary schools, universities, and currently the U.S. Air Foce (400 members) is beating out the U.S. Navy (298 members), although the U.S. Army and Marine Corps are also within the top 10 spots.
SETI@home started by simply analyzing radio signals from the Arecibo radio telescope in Puerto Rico. However, due to its success rate, the project rapidly needed to expand its scope to continue to have new information to analyze and so the Berkeley Open Infrastructure for Network Computing (or BOINC) was developed. The BOINC platform allows SETI@home to examine data from a variety of telescopes and to look at a wider radio-frequency range.
With the success of SETI@home, the development of additional distributed computing processes was inevitable and the BOINC platform has facilitated this shift. Whereas once users had to subscribe to one project at a time, BOINC allows users to designate which projects they want to be involved with and how much of their computer's time they'd like to designate to each research. One of the teams involved with the SETI project, calling themselves The Final Front Ear, is involved with 27 distributed computing projects that are based all over the world. These 175 members from 21 countries have 12 million hours completed at SETI and 26.5 million hours overall.
Among other projects, the Final Front Ear is associated with a climate prediction model (CPDN-climateprediction.net) that led to a study published in January in the journal Nature that is predicting global temperatures will rise between 3.6 and 19.8 degrees F. CPDN also divides up the work to be able to analyze a wider range of variables, including industrial carbon dioxide production trends and how heat and chemicals move between the air and ocean.
Another project a bit closer to home is Rosetta@home, run by David Baker of the University of Washington in Seattle.
Baker, a professor in the UW biochemistry department, designed Rosetta to try to help answer one of the biggest questions in science: how the sequence of amino acids in a protein relate to its three-dimensional (3D) structure and biological function. Protein structure is determined by the folding of the chain of amino acids that compose the molecule, but predicting protein structure from the amino acid sequence has remained an elusive goal.
Baker likens the problem of determining 3D structures to finding the lowest point on a planet if you were dropped at random on the surface. For example, if you landed randomly on Earth and started moving around and measuring elevation, you are unlikely to land close enough to the shore of the Dead Sea (the lowest point) to stumble across it during your exploration. Odds are pretty good that you'd be stuck in the ocean or the Andes or Sahara desert somewhere. But, if instead of landing by yourself, you bring 9,999 of your fellow researchers, the odds get much better. Similarly, Rosetta is allowing researchers to use computing time to find the lowest-energy and most likely 3D configuration of a protein based on amino acid sequences. Results compiled from Rosetta@home contributors have led to several publications in Science, but the quest continues for this "holy grail."
Perhaps the most recent figure to emerge in this distributed computing arena is Stardust@home. Unlike most processing programs which only require the user to download software and let it run on their computer, Stardust@home will be a bit more interactive. The Stardust capsule captured over a million comet particles from the comet Wild 2, and these are being tracked and analyzed by scientists in laboratories all over the country. However, in addition to those particles, the aerogel also captured interstellar dust, which is proving to be much more difficult to find and analyze.
This dust probably penetrated the aerogel to a depth of no more than 100 micrometers (the size of a dust mite) which means these particles did not leave visible tracks that would allow them to be located. To be able to find these particles (the scientists are hoping to find between 40 and 100), the scientists decided to create 1.6 million "movies” that will slowly zoom through the aerogel with viewers responsible for noting any possible sites of interest. Only if two of the three viewers that receive each movie concur about a spot of interest will the movie be reissued to a second group of viewers. If they also find possible particles, the scientists will then physically locate and examine the area to find the particle.
Because such a high level of concordance is required, Stardust@home will be the first to require contributors to go through online training and take a test before they are allowed to start analyzing data. Researchers had hoped to have Stardust@home up and running in April, but technical difficulties have set back their start date. Over 65,000 people have already signed up to analyze movies, and the training module is up and running. Prospective helpers receive instruction in identifying actual comet dust tracks, as well as how to avoid distracters like aerogel flakes, extraneous unevenness, and out of focus particles. After spending about 15 minutes learning how the videos work, participants will be able to take an exam and if they pass they will begin getting notifications when videos are available for analysis.
Stardust@home is providing a unique contribution to the whole idea of distributed computing: instead of just agreeing to let your computer work with the scientists, anyone who can pass the test actually gets to contribute to the research. Obviously this comes with some possible issues, as an untrained public may miss something or misunderstand something that they see. Regardless, these distributed projects are putting some of the fun back in the process of discovery that is science.
Amy Pletcher is pursuing a masters degree in technical communication at the University of Washington.
Image at Top:
An example of the Technicolor readings from SETI@home as it uses your computer to analyze frequencies from various parts of space. Photo: setiathome.berkeley.edu
|