Bacteria colonise various habitats in intricately composed communities, including the human gut and skin. Investigating these bacterial communities – the microbiomes – is a very difficult task. A key problem with microbiome analysis is bias, which can occur at various steps in the research of microbiomes, from the sample collection to evaluation.

Bias is particularly challenging during extraction, namely during the process of obtaining bacterial DNA. Various different extraction methods can be used to extract various types of bacterial DNA with various degrees of success, which can significantly alter the identification of microbial composition. The accuracy of scientific and clinical findings is therefore jeopardised.

Computer based method for all types of microbiome samples

Scientists at the Chair for Environmental Medicine have now developed a standardised procedure to address this problem. In a current study, Dr Luise Rauer, from Prof Avidan Neumann’s research team, who is head of environmental bioinformatics, has developed an innovative computer-based method to correct extraction bias. “The method is based on the general morphological characteristics of bacteria, so the shape of their cells and the stability of their cell walls, for example. This means we can apply it independently of the extraction method used and the specific type of bacteria. In principle, it can be used for all types of microbiome samples, so for example gut, skin, or environmental samples,” explains Rauer.

In the study, researchers compared microbiome samples from skin samples and so-called mock communities, which are artificial bacterial communities with a known composition, under various extraction methods. The new calculation method successfully reduced bias due to extraction, which led to a significantly more precise determination of the composition of the microbiome.
“Microbiome tests help us to better understand diseases. The results of this study could significantly improve microbiome analysis in clinical applications such as in diagnostics and personalised medicine. More reliable and consistent results could be decisive in clinical practice,” says Prof Claudia Traidl-Hoffmann, holder of the Chair for Environmental Medicine at the University of Augsburg and director of the university outpatient clinic for environmental medicine at University Hospital Augsburg.

The publication

Luise Rauer, Amedeo De Tomassi, Christian L. Müller, Claudia Hülpüsch, Claudia Traidl-Hoffmann, Matthias Reiger & Avidan U. Neumann. “De-biasing microbiome sequencing data: bacterial morphology-based correction of extraction bias and correlates of chimera formation.” Microbiome volume 13, Article number: 38 (2025). https://rdcu.be/d8Eun