cACP-DeepGram

cACP-DeepGram classifies anticancer peptides (ACPs) using FastText skip-gram embeddings and a deep neural network to distinguish ACPs from non-ACPs for peptide-based cancer therapy research.

Key Features:

• Deep Neural Network Integration: Employs a deep neural network (DNN) to discriminate ACPs from non-ACPs, reporting accuracies of 96.94% on training samples, 93.41% on alternate samples, and 94.02% on independent samples.
• Skip-Gram-Based Word Embedding: Represents peptide sequences using a FastText-based skip-gram embedding that captures contextual relationships within peptide sequences.
• Performance Advantage: Outperforms existing predictors by approximately 10% in prediction accuracy.

Scientific Applications:

• Academic Research: Supports analysis and classification of ACPs in peptide-based cancer research.
• Drug Discovery: Prioritizes candidate anticancer peptides for therapeutic development and identification of promising peptides.
• Targeted Therapy Development: Aids selection of peptides aimed at minimizing effects on normal cells and reducing treatment costs compared to traditional therapies.

Methodology:

Generates peptide embedding descriptors using a FastText skip-gram model and inputs these descriptors into a deep neural network classifier with optimized parameters.