Previous and current research

The human microbiome, the complex ecosystem of microorganisms colonising our body, has increasingly been recognised as an important determinant of human physiology. Modern sequencing technology and computational analysis methodology are now widely used to associate changes in microbiome composition and function with human diseases without any need to culture microbes. However, analysis and interpretation of metagenomic data remains challenging:

• Quantifying microbial (sub-)species and functions in an accurate manner consistently across various sequencing readouts (16S, shotgun metagenomics and meta-transcriptomics) is still difficult for complex communities due to the existence of many uncultured organisms.
• Microbiome data interpretation is often complicated by many factors that vary alongside the phenomenon of interest; typical confounders include differences in lifestyle, comorbidities, or treatments. Comparisons across studies (meta-analyses) are hampered by technical differences in sample handling and data generation.
• Perturbation effects on the microbiome are still poorly understood. Systematic data and predictive models on the specific effects of exposures such as drugs or dietary components on the microbiome are lacking despite this being a key aspect of personalised health and a potential entry point for designing intervention strategies for rationally modulating the microbiome.

To address these challenges, we develop software tools for high-precision profiling of both genomically characterised and unknown microbial species, and the functions encoded in their genomes and transcriptomes. To associate changes in these profiles with various host phenotypes of interest, we have investigated various statistics and machine learning tools and evaluated their applicability to microbiome sequencing data, particularly in the presence of confounding, or in cross-study comparisons. To make these validated statistical and machine-learning tools available to the community, we have published easy-to-use software pipelines. In my group, we focus on the application of this methodology to characterising the gut microbiome in colorectal cancer. For this disease, we have established diagnostic microbiome signatures that are globally validated across seven countries on three continents.

Future projects and goals

• Development and validation of statistical analysis tools suitable for multi-omics microbial community data, for longitudinal study designs, and for the detection and correction of confounders, thereby empowering large-scale cross-study comparisons (meta-analyses).
• In-depth analysis of secondary metabolism in the human gut microbiome to discover microbial factors that play a role in physiologically relevant processes as diverse as disease development (e.g. through genotoxic or pro-inflammatory metabolites), drug metabolism (through modification of the compound or its excretion/detoxification), and immune interactions (e.g. in the context of immunotherapy and organ transplantation) in collaboration with the Zimmermann, Typas, and Bork groups, as well as with clinical partners.