A robust ensemble framework for anticancer peptide classification using multi-model voting approach.

Area of Science:

• Biochemistry
• Computational Biology
• Bioinformatics

Background:

• Anticancer peptides (ACPs) show promise for cancer therapy.
• Accurate identification of ACPs is challenging due to sequence complexity and biological interactions.

Purpose of the Study:

• To develop a novel machine learning framework for enhanced anticancer peptide classification.
• To integrate multiple feature sets for improved identification accuracy.

Main Methods:

• Utilized a machine learning approach for ACP classification.
• Integrated sequence composition, physicochemical properties, and pre-trained language model embeddings.
• Evaluated classifier performance on benchmark datasets and compared against state-of-the-art methods.

Main Results:

• The proposed model achieved 75.58% accuracy, 0.8272 AUC, and 0.5119 MCC.
• Demonstrated superior performance compared to existing methods like UniDL4BioPep, ACPred-Fuse, and iACP.
• Achieved balanced sensitivity (0.7384) and specificity (0.773) for robust ACP identification.

Conclusions:

• Integrating diverse feature sets significantly enhances ACP classification accuracy.
• The developed framework facilitates more robust identification of anticancer peptides.
• This approach aids in the discovery of novel ACPs for therapeutic applications.