A1137
Title: Context-aware dimensionality reduction of microbial ecosystem dynamics
Authors: Liat Shenhav - NYU (United States) [presenting]
Abstract: Complex microbial ecosystems play an important role across many domains of life, from the female reproductive tract, through the oceans, to the plant rhizosphere. The study of these ecosystems offers great opportunities for biological discovery due to the ease of their measurement, the ability to perturb them, and their rapidly evolving nature. These same properties, however, make it difficult to extract robust and reproducible patterns from these high-dimensional and multi-scale environments. To address this, a context-aware dimensionality reduction method named Joint Compositional Tensor Factorization (Joint CTF) was developed that incorporates information from the same host across time, space and information layers (e.g., microbiome, metabolome, metatranscriptome). Joint CTF identifies robust patterns in longitudinal multi-omics data, allowing for the detection of ecosystem-wide changes associated with specific phenotypes that are reproducible across datasets. This model, designed to identify robust spatiotemporal patterns, would help us better understand the nature of the microbiome from the time of its formation and throughout life.
