A massively parallel algorithm for finding non-existing sequences in genomes
We discuss a method for producing a set of absent words in a reference genome with a guaranteed Hamming distance along all positions and additional information about the number of mismatches, their location and the position of the best match. We implemented it exploiting the massively parallelism of modern GPUs hardware: the code is available at https://bitbucket.org/mfalda/cuda_keeseek/.
bioinformatics↗