Successful translational applications of microbiome research rely on computational tools that can effectively detect microbial markers that are associated with diseases, and provide explanations to the associations. We propose to develop subtractive assembly approaches to microbiome sequencing data analysis, aiming to identity microbial genes and pathways that are associated with diseases. The advantages of using subtractive assembly approaches include: 1) they significantly reduce the complexity of the fragment assembly problem by focusing only on the potential difference (genes and genomes) at the initial (instead of the final) stage of the comparative analysis pipeline, and 2) they improve the assemblies of differential genes, which are important inputs for building predictive models for disease diagnosis and characterization of treatment efficacy. We will apply our new tools to analyzing disease-associated microbiomes including those associated with type II diabetes, liver cancer, inflammatory bowel disease (IBD) and those known to be related to the efficacy of cancer immunotherapy.
Our subtractive assembly approaches will be a timely addition to the computational tools that are central to the interpretation of microbiome-disease association, leading to a better understanding of the impacts of microbial communities on human health and diseases, and the mechanisms behind. Ultimately, these tools will be useful for translational applications of microbiome data for disease diagnosis, treatment responsiveness prediction and disease prevention.