A protein mapping method based on physicochemical properties and dimension reduction.

Area of Science:

• Bioinformatics
• Computational Biology
• Protein Sequence Analysis

Background:

• Graphical mapping of protein sequences is challenging due to the complexity of 20 amino acids and their physicochemical properties.
• Existing mapping methods often rely on limited physicochemical properties.
• Developing effective protein sequence mapping methods is an active area of research.

Purpose of the Study:

• To develop a simple and effective 2D graphical mapping method for protein sequences.
• To incorporate additional physicochemical properties for richer sequence characterization.
• To create a method that aids in understanding protein sequence similarities and evolutionary relationships.

Main Methods:

• Utilized 12 major physicochemical properties of amino acids.
• Employed Principal Component Analysis (PCA) for dimensionality reduction.
• Developed a 2D graphical mapping and extracted a 20D vector to characterize protein sequences.

Main Results:

• The proposed graphical mapping exhibits one-to-one mapping, no circuit, and good visualization.
• The mapping effectively captures more physicochemical information compared to existing methods.
• The method was successfully applied to analyze protein sequence comparisons and gene cluster analysis.

Conclusions:

• The method accurately reflects the evolutionary relationships of ND6 proteins.
• Cluster analysis of influenza A (H1N1) HA genes using this method aligns with known evolutionary facts.
• The proposed graphical mapping method demonstrates significant utility in analyzing protein sequences and their evolutionary patterns.