In Silico Peptide Design: Methods, Resources, and Role of AI.

Area of Science:

• Biochemistry and Molecular Biology
• Computational Chemistry
• Drug Discovery

Background:

• Peptides are vital in biological systems, therapeutics, and biomaterials.
• Conventional peptide design faces challenges like degradation, poor bioavailability, and inefficient synthesis/screening.
• Computational approaches offer powerful alternatives for peptide exploration and discovery.

Purpose of the Study:

• To provide a comprehensive review of computational strategies for peptide design.
• To highlight the role of specialized databases, tools, and advanced methodologies.
• To discuss the impact of machine learning and artificial intelligence in accelerating peptide discovery.

Main Methods:

• Structure-based design
• Molecular dynamics (MD) simulations
• Ligand-based approaches
• Machine learning (ML) and deep learning (DL) models
• Generative artificial intelligence (AI)

Main Results:

• Computational methods enable exploration of large chemical spaces and virtual screening of numerous peptides.
• Machine learning, deep learning, and generative AI models significantly accelerate the discovery of novel peptides.
• Specialized peptide databases and advanced computational tools are critical for effective peptide design.

Conclusions:

• Computational design, particularly AI-driven approaches, offers a transformative roadmap for overcoming limitations in peptide research.
• Key challenges include data inconsistency, model interpretability, and the need for improved force fields.
• Leveraging these computational advancements is crucial for future peptide therapeutics and biomaterials.