AI-based biodesign to develop peptide therapeutics for pneumonia-causing pathogens with experimental validation

Purpose and goal:

The project develops AI-driven peptide therapeutics for pneumonia to enable faster, cost-effective solutions against pathogens. Machine learning models will predict antimicrobial activity, allergenicity, and toxicity, while generative AI creates optimized peptides. In vitro and lung organoid assays will validate findings, bridging computational and experimental research to deepen understanding of pathogen-host interactions and therapeutic potential.

Expected results and effects:

The project aims to generate AI-designed peptides with high antimicrobial activity, low allergenicity, and minimal toxicity for pneumonia. Validated models will predict these properties, creating novel peptides effective against pathogens. In vitro and lung organoid assays will confirm results, enhancing understanding of pathogen-host interactions and peptide efficacy. This approach aims to shorten drug discovery timelines, offering cost-effective therapies for respiratory infections.

Approach and implementation:

The project consists 6 work packages. WP1 will collect data on antimicrobial peptides, guiding WP2 to develop AI/ML models predicting antimicrobial activity, allergenicity, and toxicity. In WP3, generative AI will design peptides for pneumonia pathogens. WP4 will analyze cell membrane interactions, and WP5 will focus on peptide synthesis and in vitro validation with lung organoids. Finally, WP6 will manage dissemination and patenting, ensuring transition from data collection to applications.