Abstract
Biological systems comprise various levels of organization that are responsible for controllingthe processes by which living organism interact to cohesively perform physiological functions.
Recent advancements in high-throughput gene sequencing and metagenome-assembled
genomes (MAGs) have provided unprecedented molecular detail. However, relying on a
single data type limits our understanding of biological complexity. To address this limitation,
an integrative framework that merges diverse multi-omic data sources is essential.
Genome-scale metabolic models (GEMs) are powerful tools for investigating mixed
microbial populations, combining biological knowledge with computational methods to
explore metabolic interactions among various species within communities. Understanding
these microbial interactions is crucial to addressing pressing challenges in human health
and diseases, from developing effective treatments for severe diseases to designing safe
pharmaceuticals and establishing disease prevention strategies.
In this dissertation, I investigate the interplay between multi-omic data integration,
community metabolic networks, and machine learning. By combining community genomescale
metabolic models (GEMs) with machine learning techniques, statistical analyses
and fine-tune phenotypic predictions, high accuracy can be achieved in human diseases
states prediction. Community GEMs offer crucial insights into stoichiometry and the genetic
regulation of biochemical reactions within the community interaction, while machine learning
effectively disentangles biological complexity by extracting pertinent information from data.
This integrative approach not only deepens our understanding of microbial interactions but
also paves the way for innovative therapeutic strategies and improved health outcomes.
| Date of Award | 2 Apr 2025 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Claudio Angione (Supervisor) |