Artificial Intelligence (AI) driven design of macrocycle molecules against drug resistant bacteria

Description

Problem

Pseudomonas aeruginosa (Pa) is identified as a “priority one pathogen” by the World Health Organisation (“development of novel treatments is urgently needed”) and a major pathogen that can infect the lung, particularly those functions were severely compromised, such as those with cystic fibrosis (J. Med. Microbiol. 2022, 71, 12).

Solution

The project team is developing products that can potentiate the efficacy of existing therapeutics of Pa infection in the lung. Through screening, two hit potentiating molecules were identified from a commercial library of 6,000 macrocyclic compounds. Based on preliminary cheminformatic analysis, we discovered that the two hit molecules are representative of two large clusters of screened macrocycles (~1,000 analogous), which gave rich structure-activity data set for guiding the development of these two series.

Macrocyclisation is a hot topic in medicinal chemistry, and examples of machine learning (ML)/ artificial intelligence (AI) models to design and optimise macrocyclic linkers have been recently reported (Nature Communications, 2023, 14:4552 ). The student will develop an AI driven workflow centred around generative deep-learning and predictive AI models to generate and virtually screen novel macrocycle libraries. With synthetic and biological testing capacity already in place, these models will accelerate the design-make-test-analyse cycle enabling efficient exploration of chemical space and optimisation of drug metabolism and pharmacokinetic properties. Success of this project would provide a powerful example of the application of ML/AI driven optimisation in the discovery of novel treatments against antimicrobial resistance (AMR) infections and would undoubtedly lead to high impact outputs and future potential funding from research councils and industry.

Training

The student will receive comprehensive training in modern AI/ML methodologies primarily in the research group (example publication using ML: ACIE, 2022, 61, e2021145). Enhanced training resources are freely available in the community both online and in person. Whilst the student’s focus will be primarily computational, there will be opportunities to spend time in the synthetic and biological labs developing a wider understanding of drug discovery – hence the studentship is genuinely multidisciplinary and will undoubtedly enhance the student’s employability.

Please apply by completing the online postgraduate research application form here: How to apply for a PhD - University of Liverpool 

Please ensure you include the project title and reference number CCPR108 when applying.