Related Experiment Video
Updated: Jul 2, 2025
10:13
Production and Visualization of Bacterial Spheroplasts and Protoplasts to Characterize Antimicrobial Peptide Localization
Published on: August 11, 2018
11.8K
Novel antimicrobial peptides against Cutibacterium acnes designed by deep learning
Qichang Dong1, Shaohua Wang1, Ying Miao2
1Shanghai MetaNovas Biotech Co., Ltd, Shanghai, 200120, China.
Scientific Reports
|February 24, 2024
Summary
Antibiotic resistance in Cutibacterium acnes necessitates new treatments. Researchers designed novel antimicrobial peptides (AMPs) using deep learning, identifying five potent candidates effective against C. acnes.
Area of Science:
- Microbiology
- Computational Biology
- Biotechnology
Background:
- Cutibacterium acnes (C. acnes) is increasingly resistant to antibiotics, driving the need for alternative therapies.
- Antimicrobial peptides (AMPs) represent a promising class of compounds for developing novel anti-acne treatments.
Purpose of the Study:
- To design novel peptides specifically inhibiting C. acnes using a computational approach.
- To evaluate the antimicrobial activity and selectivity of designed peptides against C. acnes.
Main Methods:
- A deep learning pipeline incorporating generators, classifiers, transfer learning, and pretrained protein embeddings was utilized.
- A phylogenetic tree was constructed to enhance training data for C. acnes inhibition.
- 42 novel linear peptides were synthesized and experimentally tested for antimicrobial properties.
Main Results:
- Five designed peptides exhibited high potency and selectivity against C. acnes.
- Minimum inhibitory concentrations (MICs) for the effective peptides ranged from 2-4 µg/mL.
- The study successfully demonstrated the potential of computational methods in rational peptide design.
Conclusions:
- The designed peptides show significant promise as anti-acne therapeutics.
- Computational approaches, including deep learning, are powerful tools for developing targeted antimicrobial peptides.
