Abstract
Abstract The project developed an Artificial Intelligence (AI)-based predictive model to estimate metabolite levels from gene expression data, addressing the data scarcity in oncological metabolomics studies. By leveraging paired metabolomics and transcriptomics datasets from public cancer repositories (including the Cancer Cell Line Encyclopedia (CCLE) and NCI-60) the study designed and validated bioinformatics pipelines to map the relationship between the transcriptome and the metabolome. This computational approach provides a comprehensive framework to identify molecular drivers of cancer and offers a cost-effective alternative to experimental metabolomics by inferring metabolic states directly from accessible RNA-sequencing data.
Results achieved
The project successfully defined a robust correlation structure between the metabolome and the transcriptome, yielding the following results: • Development of Predictive Pipelines: Creation and optimization of AI-based bioinformatics pipelines (utilizing algorithms such as CorTo) capable of predicting metabolite quantities from transcript expression profiles with an accuracy exceeding 70% for specific metabolites, such as 1-methylnicotinamide. • Validation: Successful testing and validation of predictive models on independent public datasets, including the Cancer Genome Atlas (TCGA), demonstrating the biological relevance of the approach. • Scientific Dissemination: Publication of core findings in peer-reviewed, open-access journals, notably Cavicchioli et al. (2022) [PMID: 35409231] and Beccacece et al. (2023) [PMID: 37505532], with subsequent applications in transcriptomics and metabolomics research. • Methodological Impact: Establishment of a novel computational framework that enables metabolic investigation across various cancer types, facilitating broader multi-omics integration and reducing reliance on large-scale experimental infrastructures.Project details
Unibo Team Leader: Federico Manuel Giorgi
Unibo involved Department/s:
Dipartimento di Farmacia e Biotecnologie
Coordinator:
ALMA MATER STUDIORUM - Università di Bologna(Italy)
Total Eu Contribution: Euro (EUR) 196.825,00
Total Unibo Contribution: Euro (EUR) 11.089,00
Project Duration in months: 24
Start Date:
28/09/2023
End Date:
28/02/2026
