Figure 1.

Cascade of events in the WNT Signalling Pathway. Our approach is to combine critical features (genes forming the model) from the most recent mathematical model of the WNT STN, by Kogan et al., with adjustable machine learning models [27]. The schematic represents the initial steps of activation of the WNT pathway. WNT binds to FZD (frizzled), and the co-receptor LRP5/6 (low-density lipoprotein receptor-related protein 5/6), which is recruited in the vicinity of FZD receptors upon WNT binding. This binding prevents the formation of the whole destruction complex consisting of AXIN, APC, and GSK3β (Glycogen Synthase Kinase 3 Beta), thereby causing rising levels of β-catenin, β-catenin translocation into the nucleus, and expression of WNT target genes, CCND1 (Cyclin D1), MYC, MYCN, AXIN2, and FZD7 [27]. Using machine learning tools, such as PROGENy, the WNT pathway activity can be calculated for each individual patient by analysing their gene expression [35,36]. Created with BioRender.com.