Fast alignment-free phylogeny reconstruction using spaced words

In our presentation, we give an overview about the spaced-words approach that we developed during the last years. Spaced words are words with wild-chard characters at certain positions defined by a binary pattern P of match and don’t-care positions. We show that standard word-based approaches to sequence comparison can be improved if spaced words are used instead of contiguous words. To find suitable patterns, we implemented a program called rasbhari that optimizes patterns for alignment-free sequence comparison and database searching. Most recently, we developed a Filtered Spaced Words approach that accurately estimates phylogenetic distances between genomic sequences. In this approach, we identify matching spaced words between a pair of genomic sequences and estimate distances based on the nucleotides aligned to each other at the don’t-care positions.