15 March 2012

Gamers in scientific work are more efficient than a supercomputer

Gamers beat the computer in the DNA game

<url>Humans are 70 percent of the time better than a computer at aligning several small DNA sequences.

With this result, the Phylo experiment, conducted by Canadian and American researchers since 2010, ended. The results of their work are published in the journal PLoS One (Kawrykow et al., Phylo: A Citizen Science Approach for Improving Multiple Sequence Alignment).

Scientists created a special game shell in which the player was required to align several DNA sequences in such a way that the number of matches between them was maximum. For simplicity, DNA was represented by colored squares of four colors, which corresponded to the four nucleotides of which it consists. If the player coped with the task in the allotted time, his result was entered into the Phylo database.

Further, the authors compared the obtained alignments with those that one of the most advanced MULTIZ multiple alignment programs at the moment is capable of doing. If the player's result exceeded the result of the computer algorithm, then his name was entered into the “hall of fame” of Phylo. It turned out that out of 350 thousand alignments made by three thousand players since the beginning of the experiment, 70 percent turned out to be more accurate than computer ones.

At the moment, there are algorithms for guaranteed optimal alignment of no more than two sequences. The complexity of such a problem depends on the square of their length n (complexity O (n 2)). If multiple alignment is required, then the complexity increases in proportion to the length of the sequences n, raised to the power of their number k (complexity O (n k)) and is in most cases inaccessible to modern computers at k>3. Therefore, programmers use heuristic algorithms, which, as can be seen from the success of the players, give a very approximate result.

DNA and protein sequence alignment algorithms are very important for modern biology. Proper alignment is often the main source of information about the functioning and kinship relationships of genes and organisms.

The experiment in its concept of attracting the general public to solve urgent scientific problems resembles the FoldIt project, during which volunteers fold proteins into three-dimensional structures. In turn, FoldIt was inspired by the SETI network project to search for extraterrestrial life.

Portal "Eternal youth" http://vechnayamolodost.ru
15.03.2012

Found a typo? Select it and press ctrl + enter Print version