Assembly and Application to the Tomato Genome

Abstract

The computational process of reconstructing a genome by assembling large amounts of raw sequencing data into long DNA fragments poses great challenges. This chapter illustrates current genome sequencing technologies and assembly algorithms by example of the tomato genome sequencing project. Over the last decade, “Next Generation Sequencing” technologies have placed great emphasis on efficient library preparation, high throughput and long read length. These developments have pushed the evolution of genome assembly approaches from greedy overlap-layout-consensus approaches that were used to assemble Sanger sequences, to de Bruijn graph and string graph approaches that are currently in use to assemble these new types of sequencing data produced in large volume. Nonetheless, many species still lack a high-quality, gold-standard genome sequence as genome assembly is still far from a solved problem. Several approaches have been developed to estimate the quality of assembled genome sequences and to perform so-called genome finishing, a complicated and costly procedure to complete the unresolved regions of the genome.  We expect that within this decade sequencing technologies will undergo another dramatic improvement, resulting in “Third Generation Sequencing” technologies with which chromosomes and genomes can be sequenced in their entirety with high accuracy. Plant breeding will benefit enormously from this development, providing breeders with the tools, data and understanding to design new traits and varieties from natural and induced genetic variation in an entirely rationalized and economical manner, and much beyond our current capabilities. The tomato genome described here was sequenced within an international collaboration and its completion spanned almost a decade. The novel sequencing technologies that were invented and commercialized during the course of this effort resulted in the generation of multiple types of sequence datasets. This in turn required development and application of state-of-the-art bioinformatics approaches to process the vast and varied datasets in order to produce a near-complete and high quality genome assembly.