Selecting antibacterial aptamers against the BamA protein in Pseudomonas aeruginosa by incorporating genetic algorithm to optimise computational screening method.

Antibiotic resistance is one of the biggest threats to global health resulting in an increasing number of people suffering from severe illnesses or dying due to infections that were once easily curable with antibiotics. Pseudomonas aeruginosa is a major pathogen that has rapidly developed antibiotic resistance and WHO has categorised this pathogen under the critical list. DNA aptamers can act as a potential candidate for novel antimicrobial agents. In this study, we demonstrated that an existing aptamer is able to affect the growth of P. aeruginosa. A computational screen for aptamers that could bind to a well-conserved and essential outer membrane protein, BamA in Gram-negative bacteria was conducted. Molecular docking of about 100 functional DNA aptamers with BamA protein was performed via both local and global docking approaches. Additionally, genetic algorithm analysis was carried out to rank the aptamers based on their binding affinity. The top hits of aptamers with good binding to BamA protein were synthesised to investigate their in vitro antibacterial activity. Among all aptamers, Apt31, which is known to bind to an antitumor, Daunomycin, exhibited the highest HADDOCK score and resulted in a significant (p < 0.05) reduction in P. aeruginosa growth. Apt31 also induced membrane disruption that resulted in DNA leakage. Hence, computational screening may result in the identification of aptamers that bind to the desired active site with high affinity.

The role of artificial intelligence in the battle against antimicrobial-resistant bacteria.

Antimicrobial resistance (AMR) in bacteria is a global health crisis due to the rapid emergence of multidrug-resistant bacteria and the lengthy development of new antimicrobials. In light of this, artificial intelligence in the form of machine learning has been viewed as a potential counter to delay the spread of AMR. With the aid of AI, there are possibilities to predict and identify AMR in bacteria efficiently. Furthermore, a combination of machine learning algorithms and lab testing can help to accelerate the process of discovering new antimicrobials. To date, many machine learning algorithms for antimicrobial-resistance discovery had been created and vigorously validated. Most of these algorithms produced accurate results and outperformed the traditional methods which relied on sequence comparison within a database. This mini-review will provide an updated overview of antimicrobial design workflow using the latest machine-learning antimicrobial discovery algorithms in the last 5 years. With this review, we hope to improve upon the current AMR identification and antimicrobial development techniques by introducing the use of AI into the mix, including how the algorithms could be made more effective.