How AI computational model is being utilized to battle COVID-19
In their endeavors to battle the pandemic, scholastic foundations have been staying longer than required on different creative systems. A group of scientists at the Indraprastha Institute of Information Technology (IIIT) Delhi containing Angshul Majumdar, a teacher in the Department of Electronics and Communication Engineering and Ph.D. researcher Aanchal Mongia has dealt with an AI-based computational model for medicating repurposing in the treatment of COVID-19.
How can it work?
” The AI model looks at the database of all antivirals that have been used for curing virus-borne diseases such as dengue, chikungunya, etc. and at the chemical structure of similar viruses to the coronavirus. Based on the similarities, the AI model recommends existing antivirals that may be useful in treating COVID-19,” says Majumdar. The group’s AI model has been created in a joint effort with Dr. Sanjay Kumar Saha of IPGME&R, Kolkata, and Emilie Chouzenoux, a scientist at INRIA, Saclay, Paris, France, who have helped in determining the AI calculation and sharing their clinical bits of knowledge.
Why not an immunization
While preliminaries are on to discover an immunization for forestalling the coronavirus disease, it might even now take a year or two for the antibody to be accessible for mass use. “This makes repurposing of existing drugs the best means to address COVID-19. Using clinical expertise, one would ideally need to try out 100+ existing antivirals from the database of drugs, which would be both time consuming and a drain on resources. Our AI model will help clinicians select a few antivirals that have better chances of succeeding,” says Aanchal.
Until now, the AI model has chosen medications, for example, Remdesivir, Umifenovir, Ribavirin, and Sofosbuvir as planned fixes, and which are all experiencing preliminaries for COVID-19. “This indicates that the computational model and the state of clinical practice are in sync. However, since the coronavirus is known to mutate quickly, the antiviral drugs have to evolve. In such cases, the AI model can help clinicians select only a few effective ones to treat the mutated version,” Majumdar says. His group has imparted their discoveries to clinicians and specialists at IPGME&R for clinical preliminaries. It plans to improve the computational model by utilizing profound learning methods and connecting with hereditary building specialists to comprehend the “science” behind COVID-19.
Application and exploration possibilities
“The AI computational model calls for a multidisciplinary approach involving data science, bioinformatics, chemistry, biotechnology, etc., for disease diagnosis and treatment. Owing to the fact that COVID-19 often requires a combination of drugs, the AI model can predict the right set of drugs to be used,” says Michael Gromiha, teacher, Department of Biotechnology, IIT-Madras who has been utilizing AI to investigate infection-causing transformations.
The AI model, he includes, holds massive possibilities for understudies of exploration because of the current blast in information and AI’s different applications. They can look for professions that incorporate investigation in the pharmaceutical and bioinformatics part or seek after post-doctoral cooperation in the scholarly community should the field intrigue them.