Overview of synteny prediction methods.

Existing methods for identification of conserved syntenic regions differ in many criteria like the type of the homology data used, strand awareness and gene order conservation (colinearity/clustering, whereby colinearity implies clustering). The ‘Homology type’ column indicates how matches between genes are scored and what kind of homology data is used, ‘binary 1∶1’ denotes for example best-reciprocal hits and ‘binary’ indicates that some kind of binary gene family concept like COGs, Inparanoid or EnsEMBL can be used. ‘gene-specific’ means that BLASTP similarities or conserved distances are used in the scoring function. Of all the listed approaches, OrthoCluster is the most flexible. ‘*’ For the two most recent approaches (OrthoCluster and MCMuSeC), computation of 17 vertebrate genome comparisons proved to be not feasable.