Researchers at the University of Quebec have developed high-performance gene predictors using Jacket to accelerated their MATLAB® code. This work has been published in BMC Research Notes and is freely available here.
Computerized approaches to studying the human genome are challenged by the exploding amount of data, which doubles roughly every 6 months. In order to deal with this burgeoning datasets, demands for faster processing power continue to arise.
This work focuses on predicting genes using frequency analysis with FFTs and with an equivalent technique known as Goertzel’s algorithm. In these applications, the emphasis of this paper is to propose tools to geneticists and molecular biologists for the prediction or identification of new genes using existing complementary strategies. The criteria for these tools are speed, reliability, accuracy and ease of use, thus requiring little training.
The results were impressive. At the larger sizes of sequences, Jacket is 43X faster than the Parallel Computing Toolbox® (PCT)’s PARFOR loop with 8 CPU worker threads.
The authors avoided the use of the GPU functionality available in PCT(tm), because “there are no commands that are equivalent to PARFOR that can execute multiple FOR loops in parallel on the GPU. While the command arrayfun does offer similar functionality, it removes some of the flexibility needed for this particular algorithm given our need for processing a sliding window within an array. This method was therefore not considered further.” This is yet another example that PCT(tm) gives poor performance.
Special thanks to the authors, Sylvain Robert Rivard, Jean-Gabriel Mailloux, Rachid Beguenane, and Hung Tien Bui, for their great work and sharing their insights with the rest of us. We look forward to seeing the continued developments that come out of this group!