SSAW: A new sequence similarity analysis method based on the stationary discrete wavelet transform

Area of Science:

• Bioinformatics
• Computational Biology
• Data Science

Background:

• Alignment-free sequence similarity analysis offers computational efficiency compared to alignment-based methods.
• The increasing volume of biological sequence data necessitates faster analysis techniques.

Purpose of the Study:

• To introduce a novel alignment-free sequence similarity analysis method called SSAW (Sequence Similarity Analysis using the Stationary Discrete Wavelet Transform).
• To evaluate the performance of SSAW in sequence clustering and classification tasks.

Main Methods:

• SSAW extracts k-mers from sequences and maps them to a complex number field.
• The Stationary Discrete Wavelet Transform (SDWT) is applied to convert complex number series into numerical feature vectors.
• Sequences are represented as feature vectors for downstream analysis.

Main Results:

• SSAW demonstrated competitive or superior performance against state-of-the-art alignment-free methods in accuracy, F-score, precision, and recall.
• The method achieved significantly faster running times in most tested applications.
• SSAW effectively handles large-scale sequence data for clustering and classification.

Conclusions:

• SSAW is an efficient and accurate alignment-free method for sequence similarity analysis.
• Its performance makes it suitable for modern applications dealing with vast amounts of sequence data.
• The method provides a valuable tool for bioinformatics and computational biology.