Agenda

Friday, October 18

8:00 am Continental Breakfast

8:50 Opening Remarks

Christopher Woelk (Merck Exploratory Science Center) & Curtis Huttenhower (Harvard T.H. Chan School of Public Health)

9:00 Exploring the Boundaries of Host Genetics and the Microbiome in Colon Cancer

Bonnie Hurwitz (University of Arizona)

Colorectal cancer (CRC) ranks as the third most deadly cancer with approximately ~50,000 deaths in the U.S. alone. Chronic intestinal inflammation plays a key role in CRC development given that patients with inflammatory bowel disease, ulcerative colitis or Crohn’s disease have an increased risk of CRC. At the same time, the microbiome may also contribute to inflammation in the gut. One crucial area of research is studying the interaction between host genetics and the microbiome. Here, we untangle the relationship between TGF-β, a gene involved in immune system control, H. hepaticus, a known inflammatory bacteria, and the microbiome. Using a TGFβ-signaling-deficient mouse model of human colorectal cancer (CRC), we show that the disruptions host-TGFβ-signaling produce changes within the microbiome, suggesting a causative role in the development of colon cancer. We reveal specific bacterial species that contribute to 4 major pathways known to affect human CRC: LPS production leading to inflammatory bowel, butyrate metabolism invoking the Warburg effect, polyamine biosynthesis, and oxidative phosphorylation leading to reactive oxygen and nitrogen species. With the exception of LPS production, surprisingly the gut microbiome is a major contributor to pathway dysregulation heretofore thought to originate entirely from the colon mucosal epithelium. Studies that integrate host-genetics in model systems with changes in the microbiome have the potential to identify bacterial contributions to human disease. This information can then be used to test the efficacy of pre/probiotics in disease treatment, and identify new targets based on the cross-talk between the host and microbiome.

9:30 Statistical and Network-Based Approaches for Multi-Omics Integration

Jianguo “Jeff” Xia (McGill University)

Multi-omics approaches are increasingly employed in current microbiome studies to understand compositions and functions, as well as potential interactions with the hosts. By focusing on small molecules downstream of the omics’ cascades, metabolomics provides unique and functional insights into the underlying biochemical processes that regulate these interplays. Global metabolomics based on high-resolution LC-MS is particularly appealing in this direction. However, significant challenges remain for integrative analysis of these two types of omics data. In this talk, I will present our journey on using different statistical and network-based approaches to understand multi-omics datasets from two recent microbiome studies.

10:00 COFFEE BREAK

10:30 Multi-Omics for Human Microbiome Population Studies

Curtis Huttenhower (Harvard T.H. Chan School of Public Health)

Human microbiome research at population scale has now shown that successful applications of microbiome science to health will require a surprising degree of personalization, ecological, and molecular detail. This implies both quantitative methods and study designs that can pinpoint microbial mechanisms of bioactivity, despite the microbiome’s tremendous degree of inter-individual and inter-population variability. Projects such as the second phase of the Human Microbiome Project (HMP2) have particularly highlighted this need, by identifying dynamical, longitudinal mechanisms of host-microbiome interaction in disease. I will discuss results from the HMP2 Inflammatory Bowel Disease Multi’omics Database (IBDMDB) study, its associated data and methodological resources, and the challenges remaining in effectively designing multi-omic microbiome population studies to focus on specific targets for translation and improvements in health.

11:00 Database of Microbial Host Interaction (DoMI)

Cheng Fang (Clarivate) & Julia Maritz (Merck Exploratory Science Center)

The microbiome and its connection to a myriad of human diseases has quickly opened a new research front across the scientific community. The human microbiome includes thousands of microbial strains that vary in presence and abundance across subjects, geographies, and normal and disease conditions. Over 300,000 microbial genes are expressed on average in each human gut microbiome. Despite the availability of large-scale microbiome and host data, there is a lack of methodology to integrate these data across species and understand how bacterial species in the gut interact with host cells. To drive away from species to mechanistic interpretation, a Database of Microbial-Host Interactions (DoMI) is being constructed through a proof of concept collaboration between Clarivate Analytics and Merck. Currently, 500 publications have been curated resulting in the capture of >12,500 cross species protein-protein interactions and >40 compound-protein interactions including metabolites and microbe-associated molecular patterns. This data will complement and be integrated with the Clarivate systems biology MetaBase database containing over 2 million biochemical and molecular interactions in human, mouse and rat, including over 7,000 metabolite interactions. Metagenomic sequence data from microbiome samples will be used to construct a differentially abundant gene matrix annotated with Uniprot ID’s for mapping onto DoMI. At the same time, transcriptomic data from the host (e.g. RNA-Seq) will be used to derive a gene expression matrix that may also be mapped onto DoMI so that the subnetworks related to microbe host interactions and the propagation of signaling pathways into the host may be extracted for network visualization. When this POC is complete, DoMI will drive mechanistic understanding of disease and drug action by identifying the relevant interactions between the microbiome and the host. Finally, this effort will be built out as a consortium in order to expand the interactions in DoMI with additional host-microbe as well as microbe-microbe interactions and add additional functionality.

11:45 LUNCH

12:45 pm Briefing: Breakout Panel Sessions

1:00 Breakout Panel Sessions

1. Microbiome Experimental Design (including Metagenomics, Metatranscriptomics, Metabolomics & Metaproteomics)
2. Systems Biology & Data Integration
3. Microbiome Biomarkers to Clinic
4. Microbiome Drug Discovery to Clinic

2:20 COFFEE BREAK

2:40 Feedback: Breakout Panel Sessions

4:00 KEYNOTE PRESENTATION: Population and Functional Genomics of Host-Microbiome Interactions

Ran Blekhman (University of Minnesota)

Although the microbiome is influenced by environmental factors, a strong host genetic factor is also expected to control the interaction between humans and the microbiome. Understanding the relative role of genetic and environmental factors in host-microbiome interactions is a central goal in human disease research. Research in the Blekhman lab approaches these questions by considering the microbiome as a quantitative trait, and using population genetics and machine learning to directly map host genomic factors controlling the variation in the microbiome, as well as identify individual host genes and pathways that are regulated by the microbiome. I will describe recent projects aiming to: (1) identify host genes that are associated with microbiome composition in humans; (2) quantify the heritability of the microbiome; (3) characterize how the interactions between host gene mutations and microbial taxa affect disease, with a focus on colon cancer; and (4) illustrate how variation in the microbiome affects host gene regulation.

5:00 Closing Remarks

Daria Hazuda (Merck Exploratory Science Center)

5:10 SOCIAL MIXER

6:15 Close of Day