A collaborative team of researchers from IIIT-Delhi and Inria Saclay, France has developed an artificial intelligence (AI)-driven system to identify alternative antibiotics for treating drug-resistant bacterial infections. The new tool aims to support clinical decision-making by repurposing existing antibiotics in the face of rising antimicrobial resistance (AMR).

Key Highlights

Hybrid AI approach to drug repurposing

• The system combines clinical usage data, bacterial genomic information, and chemical profiles of antibiotics.

• Instead of relying solely on fixed databases, the model detects real-world patterns in treatment effectiveness.

Global collaboration and technical leadership

• Project led by Dr. Emilie Chouzenoux (Inria Saclay) and Dr. Angshul Majumdar (IIIT-Delhi), with support from engineer Stuti Jain and graduate students Kriti Kumar and Sayantika Chatterjee.

Tested on multidrug-resistant infections

• The AI tool was tested against cases involving:

  1. Klebsiella pneumoniae (hospital-acquired infections)

  2. Neisseria gonorrhoeae (sexually transmitted diseases)

  3. Mycobacterium tuberculosis (tuberculosis)

• In each instance, the model suggested viable alternative antibiotics, which were cross-verified with resistance data and expert input.

Practical applications in low-resource settings

• The system can assist clinicians in resource-constrained hospitals by reducing treatment delays and improving antibiotic stewardship.

• It is also designed to work in public health settings lacking full diagnostic infrastructure.

Expert Commentary
 “Our method makes it possible to use existing knowledge more effectively and opens the door to smarter, faster responses to AMR,”
 – Dr. Angshul Majumdar, IIIT-DelhiThe AI-powered antibiotic repurposing system presents a promising tool to combat rising antimicrobial resistance. By using real-world clinical and molecular data, it enables faster, more efficient decision-making—especially critical in hospitals dealing with multidrug-resistant infections and limited diagnostic capabilities.