Human members of the eukaryotic protein kinase family

Abstract

Background: Eukaryotic protein kinases (EPKs) constitute one of the largest recognized protein families represented in the human genome. EPKs, which are similar to each other in sequence, structure and biochemical properties, are important players in virtually every signaling pathway involved in normal development and disease. Near completion of projects to sequence the human genome and transcriptome provide an opportunity to identify and perform sequence analysis on a nearly complete set of human EPKs.

Results: Publicly available genetic sequence data were searched for human sequences that potentially represent EPK family members. After removal of duplicates, splice variants and pseudogenes, this search yielded 510 sequences with recognizable similarity to the EPK family. Protein sequences of putative EPK catalytic domains identified in the search were aligned, and a phonogram was constructed based on the alignment. Representative sequence records in GenBank were identified, and derived information about gene mapping and nomenclature was summarized.

Conclusions: This work represents a nearly comprehensive census and early bioinformatics overview of the EPKs encoded in the human genome. Evaluation of the sequence relationships between these proteins contributes contextual information that enhances understanding of individual family members. This curation of human EPK sequences provides tools and a framework for the further characterization of this important class of enzymes.

Figure 1

Dendrogram summarizing sequence groupings and branching patterns of EPKs derived from the phenogram. The program TreeExplorer was used to edit the main phenogram (see Additional data files), collapsing branches composed entirely of sequences that cluster with one another. Branch nomenclature was guided by previously accepted subfamily nomenclature, but differs from the classical subfamily nomenclature where justified by the structure of the phenogram. Sequences that do not clearly belong to a particular cluster were left in the figure as singletons. Because no branch lengths were specified for partial sequences present in the phenogram, and because only a portion of the catalytic domain of these sequences is available for comparison, assigning partial sequences to clusters would have been subject to significant error, and was not attempted. Partial sequences are therefore not included in this figure. Subfamilies are labeled in red, while singletons are labeled in blue. The relationship between each EPK, the assigned branch names, and the classical subfamily nomenclature is presented in the EPK data table (see Additional data files).