AI-Powered Platform Enables Rapid Detection of Drug-Resistant C. Auris Pathogens

Infections caused by the pathogenic yeast Candida auris pose a significant threat to hospitalized patients, particularly those with weakened immune systems or those who have invasive medical devices. The fungus spreads easily in healthcare settings, survives for long periods on surfaces, and is often resistant to standard disinfectants and antifungal drugs. Once infections reach the bloodstream or vital organs, they can become life-threatening. Current diagnostic methods are slow, costly, and poorly suited for rapid treatment decisions. Researchers have now developed a fast, precise test that can quantify C. auris and its antifungal resistance directly from routine patient swabs.

The technology, developed by researchers specializing in CRISPR diagnostics, single-molecule detection, and computational analysis at Wyss Institute for Biologically Inspired Engineering (Boston, MA, USA), integrates CRISPR-based SHERLOCK diagnostics with ultra-sensitive single-molecule microarray technology. By combining molecular detection with real-time fluorescence readouts, the system enables rapid and quantitative analysis of fungal DNA from easily obtained swab samples.

The method, called digital SHERLOCK or dSHERLOCK, uses thousands of parallel single-molecule reactions to detect C. auris genetic material with single-nucleotide precision. Fluorescent signals generated during the reaction are monitored in real time and interpreted using machine learning algorithms. This design allows the test not only to detect the presence of C. auris but also to measure fungal load and identify resistance-causing mutations against commonly used antifungal drugs, even in mixed populations with varying susceptibility.

The researchers validated dSHERLOCK using patient surveillance swabs provided by a state public health mycology laboratory. The assay detected C. auris within 20 minutes and quantified fungal burden within 40 minutes, a major improvement over existing workflows that can take up to a week. The study, published in Nature Biomedical Engineering, showed that dSHERLOCK accurately identified mutations linked to resistance against azole and echinocandin antifungals, revealing resistance patterns that standard diagnostics often miss.

By delivering rapid, quantitative results at the point of care, dSHERLOCK could help clinicians select effective antifungal therapies sooner and reduce unnecessary drug exposure. The technology also offers a way to monitor resistance evolution during outbreaks, improving infection control in hospitals and nursing homes. The researchers plan to further adapt the platform for decentralized clinical use and explore its application to other fungal and microbial pathogens where rapid resistance profiling is critical.

“The capabilities that we are introducing with dSHERLOCK satisfy the major clinical requirements for a next-generation assay to rapidly identify and quantify the C. auris burden in easily obtained patient samples and produce a quantitative snapshot of the AMR landscape in individual samples,” said co-senior author James Collins, Ph.D., who led the effort. “This has not been possible using previous diagnostic methods and is a technological feat.”

Related Links:
Wyss Institute for Biologically Inspired Engineering