Classification and functional annotation of eukaryotic protein kinases

Area of Science:

• Bioinformatics
• Molecular Biology
• Evolutionary Biology

Background:

• Reversible protein phosphorylation is a key eukaryotic signaling mechanism.
• Accurate classification of protein kinases is crucial for understanding cellular processes.

Purpose of the Study:

• To develop a novel hidden Markov model library for precise protein kinase family classification.
• To assess the library's performance against established methods like BLASTP and Pfam.
• To apply the library for kinome annotation and evolutionary analysis across eukaryotes.

Main Methods:

• Development of a multilevel hidden Markov model library.
• Testing the library on characterized kinomes of multiple eukaryotic species.
• Comparative analysis with BLASTP and Pfam for kinase retrieval and classification.
• Application to unclassified kinases in yeast and across 21 eukaryotic genomes.

Main Results:

• The library achieved zero misclassification rate on characterized kinomes.
• It outperformed BLASTP and Pfam in kinase retrieval and family classification.
• Yeast kinome classification improved from 66.96% to 90.43%.
• Seven kinase families (AGC, CAMK, CK1, CMGC, STE, PIKK, RIO) are conserved across all analyzed eukaryotic genomes.
• Putative tyrosine kinases (TKs) were identified in plants and amoeba, suggesting early evolution.
• Tyrosine kinase-like kinases (TKLs) were found to be abundant in plants and present in various other eukaryotes.

Conclusions:

• The developed hidden Markov model library is a superior tool for protein kinase classification and kinome annotation.
• The findings provide novel insights into the early evolution and diversification of protein kinase families in eukaryotes.
• The conserved presence of seven kinase families suggests their essential role in eukaryotic life.
• The identification of TKs in plants challenges previous assumptions and points to an ancient origin.