Effective identification and differential analysis of anticancer peptides

Area of Science:

• Biochemistry
• Computational Biology
• Drug Discovery

Background:

• Anticancer peptides (ACPs) show promise as targeted cancer therapeutics with reduced toxicity.
• Current experimental validation of ACPs is limited, hindering large-scale identification.
• Developing efficient computational models is crucial for accelerating ACP discovery.

Purpose of the Study:

• To propose and validate a novel computational method, ACP_DA, for identifying anticancer peptides.
• To improve the accuracy and efficiency of ACP prediction from sequence data.

Main Methods:

• ACP_DA utilizes peptide residue composition and physicochemical properties.
• Sequential forward selection was employed to optimize feature groups, reducing overfitting and computational cost.
• A light gradient boosting machine classifier was used for model construction.

Main Results:

• ACP_DA achieved a Matthew's correlation coefficient of 0.63 and an accuracy of 0.8129 on an independent test set.
• The method demonstrated a performance enhancement of at least 2% compared to existing state-of-the-art approaches.
• The study provides accessible code and data for the ACP_DA tool.

Conclusions:

• ACP_DA is an effective computational tool for identifying anticancer peptides.
• The method has the potential to significantly contribute to the development and optimization of ACP-based cancer therapies.
• The approach offers a valuable alternative to time-consuming and expensive experimental validation methods.