Welcome to the Home Page of MLCPP
We developed a two-layer prediction framework for machine-learning-based prediction of cell-penetrating peptides (MLCPP). The first layer predicts whether a given peptide is a CPP or non-CPP, while the second layer predicts the uptake efficiency of the predicted CPPs. We employed four machine-learning-based models, along with ensemble models for a two-layer prediction framework using features computed from the peptide sequence, including amino acid composition, dipeptide composition, atomic composition, composition-transition-distribution, and physiochemical properties. Our ensemble method (combination of random forest (RF) and extra-tree classifier) outperformed the state-of-the-art predictors in CPP prediction when tested on independent datasets; in CPP uptake efficiency, our RF-based predictor was better than the state-of-the-art predictor. We anticipate that MLCPP will be a potent tool for predicting CPPs and their uptake efficiency, and may facilitate hypothesis-driven experimental designs
Enter the peptide sequences in FASTA format (Example)