In recent years, the analysis of collectively sampled and sequenced microbial genomes has emerged as one of the key technologies in the field of environmental genomics and has become widely-known under the term 'metagenomics'. The most prominent metagenome study to date is the Sargasso Sea project, which lead almost to a twofold increase of the sequenced genes stored in public databases. Many similar projects are on the way and are expected to increase the sequence information such that metagenome sequences soon will exceed those coming from whole genome sequencing projects.
In order to transform this wealth of sequence information into biological meaning, techniques are required that cluster non-overlapping sequences that originate from one or closely related species. It has been known for long that species differ regarding the occurance of short oligonucleotides, and more recently it has been shown that these intrinsic DNA signatures carry a weak phylogenetic signal.
MetaClust is a JAVA-based tool that combines different approaches to such species-specific DNA-patterns like the GC content, Dinucleotide Relative Abundances, raw counts and statistical evaluations of short oligonucleotides and chaos game representations. All of these approaches have been implemented in a modular manner which allows them to be easily modified or complemented by further methods. MetaClust stores sequences and the corresponding results computed by each of the modules in a relational database management system. Once all intrinsic signatures for a given set of sequences have been computed, similar sequences can be organized into bins by means of a clustering algorithm. While different clustering algorithms can be applied, evaluation showed that complete linkage clustering with the Euclidian distance as distance measure performed best. It could be demonstrated that this approach is feasable with artificial test sets as well as with real-world metagenome data.. In all cases, MetaClust was able to cluster metagenome fragments to organism bins with high confidence.

RibAlign

Please follow this link: http://www.megx.net/ribalign

TETRA

Please follow this link: http://www.megx.net/tetra