HMGU_Icon_Pioneer_Campus

Using AI and DNA Signals to Detect Living Microbes

AI Featured Publication Pioneer Campus

A research team at Helmholtz Munich and Helmholtz AI, led by Prof. Lara Urban, has demonstrated that artificial intelligence (AI) can be used to distinguish between living and dead microorganisms by analysing raw data signal from nanopore DNA sequencing. This technique measures changes in electrical current as DNA strands pass through nanoscale pores, generating detailed signal patterns beyond the DNA sequence alone. The researchers trained a model to infer microbial viability and applied it to a mock metagenomic dataset – a controlled mixture simulating the complexity of real-world environmental samples like soil, water, or the human body, without isolating individual organisms. The findings could improve the speed and accuracy of DNA-based diagnostics, with potential applications in public health, clinical care, and environmental monitoring.

New Possibilities in Pathogen Detection

Genomic-based diagnostics have emerged as promising alternatives to conventional culture-based methods, offering advantages in speed, accessibility, sensitivity, and precision (*). However, they have a key limitation: they cannot easily determine whether detected DNA originates from viable – and therefore potentially infectious – organisms. This uncertainty affects decisions in clinical microbiology, infection control, and environmental health monitoring.

“Nanopore sequencing gives us much richer data than most other technologies,” says Urban. “The raw signal – produced as DNA passes through the nanopores – has already been used to detect chemical modifications to DNA. We now show, for the first time, that this signal also holds information about whether microbes are alive, which is critical for many of our applications in public health genomics."

AI Models Reveal Microbial Viability

“We show that AI models can be trained to predict whether DNA comes from a dead or alive microorganism,” so Harika Ürel, doctoral researcher at Helmholtz Munich and the first author of the paper. “This is the first time such a prediction has been made using raw sequencing data.”

“Nanopore technology has already transformed genomics with its portability and data richness. What we add here is first evidence that we can infer biological states directly from the sequencing signal, which is important for real-time, genome-based pathogen monitoring,” adds Urban.

 

(*) as previously demonstrated by the Urban team at Helmholtz Munich (see Sauerborn et al., 2024, Nature Communications; and Perlas et al., 2025, Virus Evolution).

 

Original publication

Ürel et al., 2025: Nanopore- and AI-empowered microbial viability inference. GigaScience. DOI: 0.1093/gigascience/giaf100

About the scientist

Prof. Lara Urban is a Principal Investigator at Helmholtz AI, where she leads research at the intersection of artificial intelligence and genomics. She also holds a position on the professorship track at the University of Zurich, where her work contributes to One Health research – an integrated approach that explores the links between human, animal, and environmental health. Contact: lara.urbanspam prevention@helmholtz-munich.de

Funding information

The research was supported by a Helmholtz Principal Investigator Grant awarded to Lara Urban at Helmholtz AI and brought together expertise from microbiologists, computational biology, and data science. It was carried out within the interdisciplinary research environment of Helmholtz AI, which focuses on applying AI across the life sciences.

lara_urban_freigestellt
Dr. Lara Urban

PI "One Health"

View profile

Related news

lara_urbanx

Awards & Grants, Pioneer Campus,

Lara Urban is Young Scientist of the Year 2022

Helmholtz Munich scientist Lara Urban has been selected by academics as the Young Scientist of the Year 2022. The academics Young Scientist Award, which is endowed with 5,000 €, honors young scientists who have made lasting advances in their…