PTGAC Model: A machine learning approach for constructing phylogenetic tree to compare protein sequences.

Area of Science:

• Bioinformatics
• Computational Biology
• Machine Learning

Background:

• Traditional phylogenetic tree construction methods, such as the Unweighted Pair Group Method with Arithmetic Mean (UPGMA), are computationally intensive.
• Accurate phylogenetic analysis requires considering the complex chemical properties of amino acids.

Purpose of the Study:

• To develop a novel machine learning-based phylogenetic tree generation model (PTGAC) utilizing agglomerative clustering.
• To provide a computationally efficient and accurate alternative for protein sequence comparison.

Main Methods:

• Utilized Principal Component Analysis (PCA) to reduce the dimensionality of amino acid chemical properties.
• Developed a novel three-component vector descriptor based on non-central moments.
• Employed Euclidean Distance and hierarchical agglomerative clustering to construct phylogenetic trees.
• Compared PTGAC performance against UPGMA and other methods using qualitative (rationalized perception) and quantitative (symmetric distance) analyses.

Main Results:

• The proposed PTGAC model demonstrated superior performance compared to existing methods in both qualitative and quantitative analyses.
• PTGAC achieved more satisfactory phylogenetic tree construction in terms of accuracy and efficiency.
• The model effectively handles protein sequences of varying lengths, showcasing its robustness.

Conclusions:

• The machine learning-based PTGAC model offers a significant advancement in phylogenetic tree generation.
• PTGAC provides an efficient and accurate alternative for analyzing protein sequence evolution.
• The model's ability to incorporate chemical properties enhances the reliability of phylogenetic inferences.