QUEZON CITY, Philippines — Researchers from the University of the Philippines Diliman – College of Science (UPD-CS) have reached a breakthrough in the global fight against “superbugs.” , a team of chemists has developed an artificial intelligence (AI) tool designed to accelerate the discovery of new antibiotics.
The tool, named ISCAPE (Interpretable Support Vector Classifier of Antibacterial Activity of Peptides against Escherichia coli), aims to address the rising threat of Antimicrobial Resistance (AMR), which renders traditional medicines ineffective.
Developed by Remmer Salas, Dr. Portia Mahal Sabido, and Dr. Ricky Nellas from the UPD-CS Institute of Chemistry, the AI tool changes how scientists screen for life-saving molecules.
- Antimicrobial Peptides (AMPs): The tool focuses on finding “peptides”—small proteins that can kill bacteria. These are seen as the next generation of antibiotics because bacteria find it harder to develop resistance to them.
- Predictive Power: ISCAPE predicts whether a specific peptide can kill or inhibit the growth of E. coli, a common bacterium responsible for food poisoning, UTIs, and sepsis.
- SMILES Integration: The system is remarkably user-friendly; it only requires a Simplified Molecular-Input Line-Entry System (SMILES) string—a line of text representing a chemical structure—to run an evaluation.
The primary value of ISCAPE is the massive amount of time and resources it saves in the laboratory.
| Feature | Traditional Discovery | ISCAPE-Enhanced Discovery |
| Process | Physical synthesis and one-by-one testing. | Digital screening of thousands of candidates. |
| Timeframe | Months to years for early screening. | Seconds to minutes per molecule. |
| Trial & Error | High; many synthesized peptides fail. | Low; researchers focus only on high-potential “active” candidates. |
| Interpretability | Results are often “black box.” | Explains which molecular features make a peptide effective. |
“Traditionally, discovering antibacterial peptides means synthesizing many candidates and testing them one by one. This process is time-consuming… ISCAPE helps address AMR by accelerating early-stage screening through data-driven design.” — Remmer Salas, Lead Researcher
While the current model is optimized for E. coli, the UP team emphasized that the technology is highly adaptable.
- Expanding Targets: With the right datasets, ISCAPE can be retrained to target other deadly pathogens, such as Staphylococcus aureus or Salmonella.
- Open Science: In a move to support the global scientific community, the researchers have made ISCAPE publicly available.
- Web Server: Accessible via Hugging Face Spaces.
- Codebase: The training data and large-scale prediction code are hosted on GitHub.
- Peer Review: The team’s research paper was recently published in the international Journal of Molecular Graphics and Modelling.
The World Health Organization has frequently cited AMR as one of the top global public health threats. By the time this tool was launched in May 2026, global healthcare costs associated with drug-resistant infections were already projected to reach into the trillions by mid-century. ISCAPE provides a homegrown, scalable solution that moves the Philippines to the forefront of computational chemistry and drug discovery.