Application for the Cell-Cell Communication Prize
Title: Modeling microbiome-host cell to cell interaction by two level network simulation with application to precision medicine
1. Problem or question being addressed
Understanding tumor phenotype is a challenging task. Indeed, independently investigating each component is not enough to characterize its entire biology. In addition, studying the interactions and relations between the different omics (i.e. Genomics, Transcriptomics, Proteomics and Metabolomics) in cancer is crucial to stratify case subtypes or distinguish essential pathways in the disease. For example, the study of the perturbation of single genes is not enough to explain the cause of abnormal behavior in a tissue that triggers the disease.
Available evidence has shown the potential involvement of microbiota in the development of different types of human cancer, though the profile of microbial community and its function in the carcinogenesis remain unclear.
The microbiome affects tumor initiation and progression through direct effects on the tumor cells and indirectly through manipulation of the immune system. It can also determine response to cancer therapies and predict disease progression and survival. Unfortunately, the mechanism established between the host and the microbiome are unknown and difficult to study.
2. Rationale for your approach
Hermida et al.1 , in a recent published paper, have demonstrated how the combination of microbiome and transcriptome can predict accurately the prognosis and drug response.
The combination of different omics is tricky due to the lack of information about the link between them. For example, there is no information about a possible interaction between the host receptor and a related microbial metabolite.
3. Details of suggested approach
As the integration of different sources of genetic data is still challenging, modeling a multi-layer network in which each layer represents a specific omic can improve cancer characterization avoiding the loss of essential properties. The multi-layer includes two different layers: transcriptome layer containing the gene expression and metabolome layer with the microbial abundance. The nodes and edges for each layer are generated using boolean implication 2,3. Moreover, the boolean implications are used in order to predict the interlayer edges which represent the host-metabolome communication . The multi-layer is used as input for the propagation method PHENSIM4,5. Taking advantage of perturbation at the genic and microbial level, it is possible to identify essential pathways in the studied disease.
4. How it will affect the broader field
This approach that exploits PHENSIM and the multi-layer host transcriptome-microbiome allows the identification of the host - microbial communication and its influence on phenotype. Moreover, recognizing the mechanism that modulates the microbiome-transcriptome cross-talk can be exploited to potentiate the efficacy of immunotherapies and decrease their toxicity. Consequently, this new model can contribute to the advance of precision medicine and lead to powerful new discoveries.
1. Hermida, L. C., Gertz, E. M. & Ruppin, E. Predicting cancer prognosis and drug response from the tumor microbiome. Nat. Commun. 13, 1–15 (2022).
2. Sahoo, D., Dill, D. L., Tibshirani, R. & Plevritis, S. K. Extracting binary signals from microarray time-course data. Nucleic Acids Res. 35, (2007).
3. Sahoo, D., Dill, D. L., Gentles, A. J., Tibshirani, R. & Plevritis, S. K. Boolean implication networks derived from large scale, whole genome microarray datasets. Genome Biol. 9, 1–17 (2008).
4. Alaimo, S. et al. PHENSIM: Phenotype Simulator. PLoS Comput. Biol. 17, (2021).
5. Maria, N. I. et al. Rapid Identification of Druggable Targets and the Power of the PHENotype SIMulator for Effective Drug Repurposing in COVID-19. Research Square doi:10.21203/rs.3.rs-287183/v1.