Similarity/Dissimilarity analysis of protein sequences based on a new spectrum-like graphical representation

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Sequence comparison is fundamental in bioinformatics for understanding evolutionary relationships.
• Existing methods may have limitations in capturing complex sequence characteristics.

Purpose of the Study:

• To develop a novel 2D spectrum-like graphical representation for protein sequences.
• To establish a new method for protein sequence similarity analysis based on hydrophobicity.
• To evaluate the utility of the proposed method on mitochondrial genome sequences.

Main Methods:

• Utilizing the hydrophobicity scale of amino acids to create a 2D spectrum-like graph.
• Characterizing the graph using frequencies of amplitudes of 4-subsequences.
• Representing protein sequences as a 17-dimensional vector descriptor.
• Performing a chi-squared (χ(2)) compatibility test.
• Applying the method to protein sequences from the mitochondrial genomes of 20 species.

Main Results:

• A novel 17D vector descriptor for protein sequences was developed.
• The method demonstrated effectiveness in analyzing similarity among mitochondrial protein sequences.
• The new approach showed utility when compared to the established ClustalW method.

Conclusions:

• The proposed 2D graphical representation and vector descriptor offer a new perspective for protein sequence analysis.
• This method provides a valuable tool for studying evolutionary relationships and sequence similarity.
• The approach is particularly useful for analyzing sequences encoded by mitochondrial genomes.