KELM-CPPpred: Kernel Extreme Learning Machine Based Prediction Model for Cell-Penetrating Peptides.

Area of Science:

• Biochemistry
• Computational Biology
• Drug Delivery

Background:

• Cell-penetrating peptides (CPPs) are crucial for transporting molecules into cells.
• Identifying novel CPPs is vital for understanding their activity.
• Wet-lab identification of CPPs is time-consuming and resource-intensive.

Purpose of the Study:

• To develop an efficient computational tool for predicting cell-penetrating peptides (CPPs).
• To accelerate the discovery of unique CPPs for potential therapeutic applications.
• To provide a freely accessible web interface for CPP prediction.

Main Methods:

• Developed a Kernel Extreme Learning Machine (KELM) based prediction model, KELM-CPPpred.
• Utilized a dataset of 408 CPPs and 408 non-CPPs.
• Input features included amino acid composition, dipeptide composition, pseudo amino acid composition, and motif-based hybrid features.
• Validated the model using an independent dataset and compared performance against SVM, RF, and ANN models.

Main Results:

• KELM-CPPpred demonstrated superior prediction accuracy compared to existing SVM, RF, and ANN approaches.
• The model effectively identified unique cell-penetrating peptides.
• A web interface for KELM-CPPpred was created and made publicly available.

Conclusions:

• KELM-CPPpred is an efficient and accurate tool for identifying cell-penetrating peptides.
• The developed model can significantly reduce the cost and time associated with CPP discovery.
• The accessible web interface facilitates broader research in CPP-mediated drug delivery.