Delhi institute's project attempting to use AI to speed up search for pneumonia therapeutics

New Delhi, Mar 31 (PTI) A new India-Sweden research collaboration at Indraprastha Institute of Information Technology, Delhi is betting on artificial intelligence helping scientists outpace deadly pneumonia germs.
     They are doing this by using AI to design peptide-based therapeutics that could one day shorten drug discovery timelines and open new pathways against hard-to-treat infections.
     Titled “AI-based bio-design for developing peptide therapeutics against pneumonia-causing pathogens with experimental validation”, the project brings together researchers from India and Sweden to use artificial intelligence and computational biology in the search for new treatments for respiratory infections.
     The research focuses on how AI can be used to help discover new medicines faster than traditional methods.
     "The traditional experimental methods are very time-consuming, invasive as well as destructive, and resource-intensive," said Professor at IIIT-Delhi's Department of Computational Biology and principal investigator of the research Docent N Arul Murugan.
     "The number of possible chemical compounds and peptides that scientists can test for a potential drug runs into billions, which is far more than can be practically screened using current laboratory methods."
     "However, screening these compounds using virtual screening approaches can be less time-consuming and more affordable, as most of the computational work can be performed on supercomputers with enormous numbers of CPUs and GPUs," he said.
     The scoring function for ranking chemicals or peptides needs to be accurate, and many AI-based approaches help us achieve reliable scoring, Murugan explained.
     The aim of the research, Murugan further explained, is to focus on designing peptide therapeutics, a class of molecules that can be tailored to target disease-causing organisms with greater specificity, especially for pathogens that cause pneumonia.
     Researchers associated with the project said the project would use machine learning, deep learning and generative AI to predict key peptide properties such as antimicrobial activity, allergenicity and toxicity, and then propose novel candidates for further testing.
     The shortlisted peptide molecules will subsequently undergo in-cell assays, in vitro studies and lung organoid-based studies to assess their real-world efficacy and safety.
     Explaining the significance of peptide therapeutics, the researchers said these molecules sit between conventional small-molecule drugs and biologics, and may offer better target specificity with potentially fewer side effects.
     The project also comes at a time when scientists worldwide are increasingly looking for alternatives to conventional anti-infective approaches amid rising concerns over drug resistance and the emergence of difficult-to-treat pathogens.
     Murugan said that if successful, the project could help shorten the long and expensive path of drug development, while also creating leads that may later be taken forward for patents and clinical development by pharmaceutical companies.
     "We are confident that by three years from now, we will have identified some lead peptide candidates against pneumonia-causing pathogens, and some of the experimental validation to establish this effect will have been carried out," he added.
     The project has received Rs 52.2 lakh from the Indian side and 2.8 million SEK from Sweden, with support also earmarked for faculty and student mobility between the two countries.
     At IIIT-Delhi, Murugan will lead the project as principal investigator, with professor G P S Raghava and Vibhor Kumar as co-principal investigators. PTI AHD










AHD SGV SKY
SKY

(This story has not been edited by THE WEEK and is auto-generated from PTI)