A horde of coding geeks has overtaken outdated algorithms for immunology studies, authoring software for analyzing immune-related genes faster than previous systems, Nature reported on a news blog. Their output made a crowdsourcing exercise from Harvard medical and business researchers an unqualified success.
The Harvard researchers tapped TopCoder.com, which hosts programming competitions, to offer just $6,000 in prize money for groups that could develop software for analyzing the genes responsible for creating antibodies and T-cell receptors in response to pathogens. Existing algorithms have been poorly accessible and slow, Nature reported. And many of the entries from the online competition blew away the old algorithms and are now available online for free.
The cool part: None of the 122 entrants work as computational biologists. By reducing some of the scientific elements into universally understood variables, the competition created an equal playing field for nonbiology buffs to race ahead with their entries. As Nature reported, strands of DNA were presented as strings rather than amalgams of genomic building blocks. The crew from Harvard reported the results in Nature Biotechnology.
"Expressing the problem in a way that computer scientists could understand was a key element in the success of this work," said Alex Bateman, a computational biologist at the European Bioinformatics Institute near Cambridge, U.K., as quoted by Nature. "I think that this article should motivate many more groups to consider this kind of outsourcing."
As previous online competitions with Foldit and other platforms have shown, crowdsourcing can be like outsourcing a problem to eager solvers who might not even be in it for the money. The Nature article dubbed this "mercenary" coding, as there was prize money at stake, but it seems like the contestants set out into new territory for very little financial gain and came back with stellar results. Perhaps they should be called missionary coders.
- check out the blog's article