Dot-plot comparisons by multivariate analysis (DOCMA): a tool for classifying protein sequences

Area of Science:

• Bioinformatics
• Computational Biology
• Protein Sequence Analysis

Background:

• Traditional protein sequence classification often relies on pairwise alignments, which can be computationally intensive and less effective for distantly related sequences.
• Developing alignment-free methods is crucial for efficient and accurate protein family analysis.

Purpose of the Study:

• To introduce a novel alignment-free method called DOCMA (Dot-plot Comparisons by Multivariate Analysis) for protein sequence classification.
• To demonstrate the effectiveness of DOCMA in delineating subgroups within protein families.

Main Methods:

• DOCMA utilizes multivariate analysis of pairwise dot-plots between sequences.
• Dot-plots are simplified by projecting diagonal similarity segments onto axes, forming a data matrix.
• A chi-squared analysis transforms the data matrix into a distance matrix, enabling sequence placement in an orthonormal Euclidean space.
• Dynamic clustering and strong cluster searching are employed for final classification.

Main Results:

• The DOCMA method was applied to protein families including globins, cytochromes c, and aminoacyl-tRNA synthetases.
• The method successfully delineated subgroups within these protein families.
• DOCMA proved effective even for classifying distantly related sequences.

Conclusions:

• DOCMA offers an effective alignment-free approach for protein sequence classification.
• The method demonstrates utility in identifying subgroups within diverse protein families.
• DOCMA provides a valuable tool for computational biology and bioinformatics research.