PSSMCOOL: a comprehensive R package for generating evolutionary-based descriptors of protein sequences from PSSM...

Area of Science:

• Computational Biology
• Bioinformatics
• Machine Learning

Background:

• Position-Specific Scoring Matrices (PSSMs) represent protein evolutionary history.
• PSSM-based features enhance predictions for protein attributes like function and localization.
• Existing methods lack an integrated tool for extracting these evolutionary descriptors.

Purpose of the Study:

• To develop a comprehensive and integrated R package for extracting PSSM-based feature vectors.
• To provide a practical tool for machine learning applications in computational biology.

Main Methods:

• Development of PSSMCOOL, an R package.
• Implementation of algorithms to generate 38 distinct PSSM-based feature vectors.
• Focus on creating a flexible and standalone tool.

Main Results:

• PSSMCOOL generates 38 PSSM-based feature vectors.
• It is the first PSSM-based feature extraction tool implemented in R.
• The package offers a comprehensive suite of evolutionary descriptors.

Conclusions:

• PSSMCOOL addresses the need for an integrated tool for PSSM feature extraction.
• The package is expected to be valuable for machine learning predictions in computational biology.
• It facilitates the utilization of evolutionary information in protein attribute prediction.