Direct carbapenemase typing from disc diffusion antibiograms with MALCA (MAchine Learning CArbapenemase)

2026-05-11 · Source: Machine learning : nature.com subject feeds · Field: Health & Wellbeing — Health & Medical Research, Clinical Care & Medical Practice, Medical Devices & Health Technology · Depth: Expert, medium

Summary

MALCA (MAchine Learning CArbapenemase) is a new machine-learning classifier designed to detect carbapenemase-producing *Enterobacterales* (CPE) and identify specific carbapenemase types directly from routine disc diffusion antibiogram results. Developed from 11,992 clinical isolates, MALCA offers two versions: MALCA-22, using 22 antibiotics, and MALCA-8, using 8 antibiotics. In an external validation study involving 8514 isolates, both classifiers demonstrated high accuracy, achieving sensitivity and specificity exceeding 96% for CPE detection, outperforming existing European and French screening algorithms. For prevalent carbapenemases like OXA-48-like, NDM, and KPC producers, MALCA achieved sensitivities over 97% and specificities above 98%. This tool provides a rapid, inexpensive diagnostic method without requiring additional reagents or human resources, facilitating earlier targeted therapy.

Key takeaway

For clinical microbiologists and infectious disease specialists managing carbapenemase-producing *Enterobacterales* (CPE), MALCA offers a significant advancement by providing rapid and accurate carbapenemase typing from standard antibiograms. You should consider implementing MALCA to enable earlier targeted therapy decisions, potentially improving patient outcomes and optimizing antibiotic stewardship without incurring extra costs for specialized confirmatory tests or additional staff.

Key insights

MALCA uses machine learning on antibiogram data for rapid, cost-effective detection and typing of carbapenemase-producing *Enterobacterales*.

Principles

• Routine antibiogram data can predict complex bacterial resistance.
• Machine learning can enhance diagnostic accuracy for bacterial typing.

Method

A stepwise random-forest pipeline processes disc diffusion antibiogram results to classify CPE and identify carbapenemase types, available in 22-antibiotic (MALCA-22) or 8-antibiotic (MALCA-8) panels.

In practice

• Integrate MALCA into existing microbiology lab workflows.
• Utilize MALCA-8 for resource-limited settings.

Topics

• Carbapenemase Typing
• Machine Learning Classifier
• Disc Diffusion Antibiograms
• Antimicrobial Resistance
• CPE Detection

Best for: Research Scientist, AI Scientist, Data Scientist

Related on AIssential

• Benchmarking Machine Learning Architectures for Antimicrobial Stewardship in Pediatric ICUs — Machine learning performance for pediatric antimicrobial stewardship is driven more by target design and data characteristics than model complexity.
• Convolutional neural networks quantify antibiotic resistance in Mycobacterium tuberculosis with diagnostic grade accuracy and predict treatment response — CNNs can predict antibiotic MICs from genomic data with clinical accuracy, leveraging diverse biological features.
• Tiny Graphene Drums Let Doctors Identify Bacteria by Sound — Graphene nanodrums can identify bacterial species and antibiotic resistance by analyzing their unique mechanical vibrations.
• Interpretable machine learning and signal processing for automated reading and quality control of lateral flow tests for schistosomiasis — An automated pipeline combining deep learning and signal processing accurately classifies schistosomiasis lateral flow tests.
• Automated Detection of Dosing Errors in Clinical Trial Narratives: A Multi-Modal Feature Engineering Approach with LightGBM — Combining sparse lexical features with dense semantic embeddings improves clinical dosing error detection.
• sebis at CRF Filling 2026: A Two-Stage Local LLM Pipeline for Medical CRF Filling — A two-stage local LLM pipeline using MedGemma-27B achieves competitive, privacy-preserving medical CRF filling with strict evidence adherence.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine learning : nature.com subject feeds.