Boosting predictive models and augmenting patient data with relevant genomic and pathway information

Samuele Buosi, Mohan Timilsina, Maria Torrente, Mariano Provencio, Dirk Fey, Vít Novácek

Abstract: Highlights•Leveraging genomics enriches graph data, enhancing cancer relapse prediction models.•Pathway imputations generated from publicly available data, improved model inference.•Graph-trained ML models offer valuable support in diverse applications as auxiliary predictive tools.•Different pathway imputation combinations influence model performance outcomes.