Fast Spectral Estimation of Genetic Homology

Technical Brief- 32 Core Testing Plan Contest
Imagine if you want to view homology of all known genetic sequences at once. As one single image it would be huge indeed. To do this we could identify pixels at positions (i,j) of the image with bright red points when the estimated similarity value of corresponding i-th and j-th DNA segments is above a predefined threshold value; green points would denote inverted identities, and so on. The questions remain: how could such a huge image be generated quickly, how would a user browse over its broad plane, and how would they pull out interesting subsequences separately for additional exploration?

There are downloads available under the Creative Commons License license. Download Now

Include in RSS: 

For more complete information about compiler optimizations, see our Optimization Notice.