Machine learning the microbiome for identification of novel antimicrobial peptides

Abuse and overuse of antibiotics has led to antimicrobial resistance (AMR), which is a major global health concern.

In the EU, there are around 30,000 AMR-related deaths per year, and the incidence of AMR is on the rise; experts have predicted a loss of 10 million lives by 2050.

One of the reasons to find alternatives to tackle this grave and complicated health challenge is that only very few antibiotics are discovered and those in the pipeline do not have the ability to circumvent AMR development in microbes. Thus, there is a need for discovery of novel antimicrobial agents that can take the place of conventional antibiotics.

In comparison to conventional antibiotics, antimicrobial peptides (AMPs) especially bacteriocins, are effective in rapidly eradicating microbes, thereby reducing AMR-related issues. Hence, the proposed project aims to identify ideal AMPs that can eradicate pathogens and reduce the emergence of AMR. The approach will be to mine the sequences of the most dominant commensal genomes to identify suitable candidate AMPs.

The identified AMPs will be stored in an online catalogue which can be leveraged by researchers working in the same field.

Furthermore, the anti-pathogenic ability of the selected AMPs will be confirmed through in-vitro techniques followed by studying their AMR developing propensity. Additionally, genes leading to AMR will be identified and knocked out to study the effect on AMR. Furthermore, the synergistic activity of the identified AMPs along with antibiotics will also be studied.

These objectives will be achieved by using advanced computational and machine learning techniques followed by in-vitro techniques.

The project will lead to discovery of AMPs with least AMR developing propensity and will also provide information about putative AMPs that can be developed along with Toscana Life Sciences (TLS) Foundation (industrial placement) to explore their therapeutic potential.