A new variant of dynamic programming for optimal and near-optimal similarity discovery in genomic sequences based on distance series Conference Paper