Distributed and paged suffix trees for large genetic databases

Abstract

We present two new variants of the suffix tree which allow much larger genome sequence databases to be handled efficiently. The method is based on a new linear time construction algorithm for "sparse" suffix trees, which are subtrees of the whole suffix tree. The new data structures are called the paged suffix tree (PST) and the distributed suffix tree (DST). Both tackle the memory bottleneck by constructing subtrees of the full suffix tree independently and are designed for single processor and distributed memory parallel computing environments (e.g. Beowulf clusters), respectively. The standard operations on suffix trees of biological importance are shown to be easily translatable to these new data structures. While none of these operations on the DST require interprocess communication, many have optimal expected parallel running times. © Springer-Verlag Berlin Heidelberg 2003.