Comparison of phosphorylation patterns across eukaryotes by discriminative N-gram analysis

Area of Science:

• Biochemistry and Molecular Biology
• Evolutionary Biology
• Bioinformatics

Background:

• The relationship between protein phosphorylation and evolutionary divergence across kingdoms/phyla remains largely unexplored.
• Amino acid sequences surrounding phosphorylation sites significantly influence protein kinase-substrate interactions.
• Traditional motif analysis methods are inadequate for large-scale comparative studies due to dataset imbalances and unique motif generation.

Purpose of the Study:

• To identify discriminative n-grams specific to different kingdoms/phyla.
• To develop a classification signature for grouping species based on phosphorylation site motifs.
• To investigate the evolutionary significance of these discriminative n-grams and their association with protein disorder.

Main Methods:

• Identification of discriminative n-grams across five species from distinct kingdoms/phyla.
• Creation of a signature comprising 5540 discriminative n-grams for inter-species classification.
• Application of classification methods to a test dataset to validate the signature's accuracy.
• Ortholog protein analysis to distinguish phosphosite-specific n-grams from species-specific inventory.
• Clustering of motifs based on physicochemical properties and secondary structure prediction.

Main Results:

• A signature of 5540 discriminative n-grams was established, capable of classifying species within their respective kingdoms/phyla.
• Distinct physicochemical patterns were observed in n-grams, with animal-specific motifs rich in basic amino acids and plant-specific motifs rich in acidic amino acids.
• Discriminative n-grams predominantly occur in disordered protein regions, exhibiting a higher percentage of random coil structure compared to overall phosphoproteins.
• Ortholog analysis confirmed that detected n-grams are properties of phosphosites, not solely due to species-specific protein content.

Conclusions:

• Discriminative n-grams effectively classify organisms into their corresponding kingdoms/phyla.
• Observed differences in n-gram patterns among species suggest group-specific kinome variations.
• These rapidly evolving, disordered n-gram regions likely contribute to evolutionary divergence across different taxa.