Fig. 1

Overview of the material, model and CNI capture

Allele specific copy-number values per gene and the gene expression profiles across different samples are used as input in the model. A nonlinear Poisson model is implemented to quantify the CNI over gene expression for each individual gene, and the results are pooled statistically to obtain a pathway level CNI. Afterwards, the CNI and absolute CNAs are used to create a landscape of CNA functionality at pathway level. A thresholding method is applied to cluster the landscape resulting into CN/non-CN aberrant and CN/non-CN-driven pathways simultaneously. The pathway level ssGSEA scores are computed using gene expression of CN-driven pathways enriched in poor and good responding patients. A survival analysis reveals prognostic pathways for all DECIDER cohort patients at diagnosis