Using, C/C++ and CUDA (leveraging GPU technology for acceleration) I accelerated a micro-array data-mining algorithm. I presented a conference paper on the project at BIOCOMP 2012

The algorithm scans microarray data profiles(of genes and transcription factors TFs)) looking for TFs that may join and act as a composite TF for the gene. The algorithm scans different combinations of TFs and genes and scores them. Based on the scores, a ranking is carried out to find candidate true-positives. The structure of the algorithm lent itself very nicely to be implemented and accelerated with CUDA. A small conference paper/presentation was made at BIOCOMP 2012, July, 2012.

A link to the paper is here.
A link to the slides is here.
A link to the source(pdf) is here. (critical codes are highlighted in yellow and green).
A link to a txt version of source here.