MLCPP 2.0: An Updated Cell-penetrating Peptides and Their Uptake Efficiency Predictor.

Area of Science:

• Biochemistry
• Bioinformatics
• Computational Biology

Background:

• Cell-penetrating peptides (CPPs) are crucial for delivering biologically active molecules into cells, offering broad biomedical applications.
• Existing machine learning predictors primarily focus on CPP identification, often neglecting uptake efficiency.
• Previous work introduced MLCPP, a predictor for CPPs and their uptake efficiency, necessitating further refinement for practical use.

Purpose of the Study:

• To develop an advanced, interpretable machine learning model, MLCPP 2.0, for simultaneous prediction of CPPs and their uptake efficiency.
• To enhance the accuracy and practical applicability of CPP prediction tools.
• To facilitate the discovery of novel CPPs for experimental design.

Main Methods:

• An updated benchmarking dataset was utilized.
• Seventeen sequence-based feature encoding algorithms and seven machine learning classifiers were explored.
• A stacking ensemble approach was employed, combining 119 baseline models through 10-fold cross-validation.

Main Results:

• MLCPP 2.0 demonstrated outstanding performance on an independent test set.
• The model significantly outperformed existing state-of-the-art CPP predictors.
• Optimal feature sets and classifiers were systematically identified for predicting CPPs and uptake efficiency.

Conclusions:

• MLCPP 2.0 offers a significant advancement in predicting both CPP identity and uptake efficiency.
• The model's enhanced performance is expected to accelerate the discovery of novel CPPs.
• MLCPP 2.0 is freely accessible, supporting hypothesis-driven research in drug delivery and molecular transport.