A first step in identifying microbiome candidates that influence a clinical parameter, production yields or any parameter of interest, is often to do a microbiome-wide association study.
We have extensive experience in setting up and analysing Microbiome-Wide Association Studies and offer tailored solutions to our clients. The challenge with microbiome-wide association studies is often the complexity and redundancy of the microbiome, as well as the complexity and/or uncertainty of the clinical parameters measured. To address this, we offer a range of systems biological solutions and offer assistance for optimal study design. All our systems biology solutions are tailored specifically to our clients based on the following key principles below.
Reduction of complexity
Most data types including microbiome, metabolomics and often clinical data, can be reduced in complexity with minimal loss of precision. For example, shotgun metagenomics data can be reduced from millions of genes to metagenomic species’ (see article) and gene modules, like plasmids, phages etc. Similarly, metabolomics data can be reduced to metabolite co-abundance modules, such that the complexity of the initial data becomes of a manageable complexity that, in turn, increases statistical power, and often increases conceptualization of the data (i.e. to species, pathways, diseases, etc.).
The functional overlapping species concept
– A way to overcome the functional redundancy and complexity of the microbiome
Most microbiomes consist of many microbial species with overlapping functional niches, and across many instances e.g. with individuals, different species may fill a given niche (see figure). Untamed, this property of microbiomes is an obstacle to classical microbiome-wide association studies. However, by grouping the species that have potential to fill equivalent ecological or functional niches, this challenge can be overcome. Furthermore, group association of functional overlapping species is often much more informative than association to individual species, as the shared properties of the species-group may give hints towards underlying mechanisms behind the association.
Figure: The association between the occurrence of species and a clinical parameter is strengthened when species that share a critical property are grouped
Multi-omics data Integration
Integration of multiple omics data - like microbiome, metabolite and clinical data - can generate critical insight to biomedical mechanistic relationships. Data integration typically includes complexity reduction, functional annotation and association linking, and almost always requires customization to the specific customer needs. Clinical-Microbiomics has leading experience with data integration of multiple different omics types, including clinical, microbiome and metabolomics (Nature, 2016); metatranscriptomics and metagenomics (Nature Microbiology, 2016), as well as intervention studies with clinical and microbiome data. On top of all this, we also offer tailored multi-omics data integration and analyses.
Figure: Overview of the workflow integrating human phenotypes, fasting serum metabolome and gut microbiome data (Nature, 2016)
With our ultra-high-resolution microbiomics protocol we can profile a microbiome down to single nucleotide variations (SNVs). For example, we have called 1.4 billion SNVs across 766 human gut microbiome samples. Combined with our metagenomic species binning protocol we can de-novo extract population structures and identify novel strains from shotgun metagenomics, resulting in an unmatched microbiome resolution that facilitates studies of real-time evolution, strain level association studies, and tracking of clonal populations.