Nanopore- and AI-empowered metagenomic viability inference

Harika Urel,Sabrina Benassou,Tim Reska,Hanna Marti,Enrique Rayo,Edward J Martin,Michael Schloter,James M Ferguson,Stefan Kesselheim,Nicole Borel,Lara Urban

DOI: https://doi.org/10.1101/2024.06.10.598221

2024-06-11

Abstract:The ability to differentiate between viable and dead microorganisms in metagenomic samples is crucial for various microbial inferences, ranging from assessing ecosystem functions of environmental microbiomes to inferring the virulence of potential pathogens. While established viability-resolved metagenomic approaches are labor-intensive as well as biased and lacking in sensitivity, we here introduce a new fully computational framework that leverages nanopore sequencing technology to assess microbial viability directly from freely available nanopore signal data. Our approach utilizes deep neural networks to learn features from such raw nanopore signal data that can distinguish DNA from viable and dead microorganisms in a controlled experimental setting. The application of explainable AI tools then allows us to robustly pinpoint the signal patterns in the nanopore raw data that allow the model to make viability predictions at high accuracy. Using the model predictions as well as efficient explainable AI-based rules, we show that our framework can be leveraged in a real-world application to estimate the viability of pathogenic Chlamydia, where traditional culture-based methods suffer from inherently high false negative rates. This application shows that our viability model captures predictive patterns in the nanopore signal that can in principle be utilized to predict viability across taxonomic boundaries and indendent of the killing method used to induce bacterial cell death. While the generalizability of our computational framework needs to be assessed in more detail, we here demonstrate for the first time the potential of analyzing freely available nanopore signal data to infer the viability of microorganisms, with many applications in environmental, veterinary, and clinical settings.

Bioinformatics

What problem does this paper attempt to address?

This paper introduces a new computational framework that utilizes nanopore sequencing technology and artificial intelligence (AI) to directly infer the microbial viability in macrogenomic samples from raw signal data. Traditional metagenomic methods cannot distinguish between living and dead microorganisms, which is crucial for assessing the functionality of environmental microbial communities, the virulence of potential pathogens, and the overall impact of the microbial community on the environment.

The research team developed a deep neural network that learns features from nanopore signal data to differentiate DNA from living and dead microorganisms. Through interpretable AI tools, they were able to identify the signal patterns on which the model relies for high-accuracy predictions of viability. In practical applications, this approach was used to estimate the viability of the pathogen Chlamydia trachomatis, which has a high false-negative rate using traditional cultivation methods.

The paper shows that their model can predict viability on different classification boundaries and is independent of methods inducing bacterial death. Although further validation of the framework's universality is needed, the research demonstrates for the first time the potential of analyzing publicly available nanopore signal data to infer microbial viability, which has various application prospects in environmental, veterinary, and clinical settings.