Seeking old wisdoms for new AMP discovery.

Area of Science:

• Biochemistry
• Genomics
• Computational Biology

Background:

• Proteolytic sites are crucial for peptide generation.
• Discovering novel peptides with therapeutic potential is an ongoing challenge.

Purpose of the Study:

• To develop a deep-learning model for predicting proteolytic sites.
• To identify antimicrobial peptides (AMPs) in modern and extinct human proteomes.
• To explore the therapeutic potential of de-extincted peptides.

Main Methods:

• A deep-learning model was constructed for proteolytic site prediction.
• The model was applied to analyze modern and extinct human proteomes.
• Bioinformatic analyses were performed to identify and characterize peptides.

Main Results:

• The study successfully predicted proteolytic sites using deep learning.
• Novel antimicrobial peptides were identified in both modern and extinct human proteomes.
• The de-extinction of archaic peptides yielded promising functional peptide candidates.

Conclusions:

• Deep learning offers a powerful tool for discovering antimicrobial peptides.
• Ancient human proteomes harbor undiscovered peptides with therapeutic potential.
• This approach revolutionizes the discovery of functional peptides for medical applications.