Create Trees
For a given set of aligned sequences (see Create an alignment) it is possible to infer their evolutionary relationships. In CLC Genomics Workbench this may be done either by using a distance based method or by using maximum likelihood (ML) estimation which is a statistical approach (see Bioinformatics explained). Both approaches generate a phylogenetic tree.
Three tools are available for generating phylogenetic trees:
- K-mer Based Tree Construction (
) Is a distance-based method that can create trees based on multiple single sequences. K-mers are used to compute distance matrices for distance-based phylogenetic reconstruction tools such as neighbor joining and UPGMA (see Distance-based methods). This method is less precise than the "Create Tree" tool but it can cope with a very large number of long sequences as it does not require a multiple alignment.
- Create Tree (
) Is a tool that uses distance estimates computed from multiple alignments to create trees. The user can select whether to use Jukes-Cantor distance correction or Kimura distance correction (Kimura 80 for nucleotides/Kimura protein for proteins) in combination with either the neighbor joining or UPGMA method (see Distance-based methods).
- Maximum Likelihood Phylogeny (
) The most advanced and time consuming method of the three mentioned. The maximum likelihood tree estimation is performed under the assumption of one of five substitution models: the Jukes-Cantor, the Kimura 80, the HKY and the GTR (also known as the REV model) models (see #a#>ximum Likelihood Phylogenysec:maximumlikelihood for further information about the models). Prior to using the "Maximum Likelihood Phylogeny" tool for creating a phylogenetic tree it is recommended to run the #o#> del Testingsec:modeltesting tool in order to identify the best suitable models for creating a tree.
Subsections
- K-mer Based Tree Construction
- Create tree
- Model Testing
- Maximum Likelihood Phylogeny
- Bioinformatics explained