Fig. 1: Overview of iPANDDA.

iPANDDA is a computational pipeline designed to identify candidate drugs and therapeutic pathways for diseases of interest. For SOX2-dependent LUSC, multi-modal datasets were integrated to construct a disease-specific protein-protein interaction (PPI) network. Data sources included transcriptional signatures (1636 genes identified from LUSC-specific TCGA analysis, 2312 DEGs from a bespoke SOX2-dependent model using RNA-seq and ChIP-seq), transcription factors (12 from ENCODE), LUSC dependency data (157 from Open Targets), and SOX2-interacting proteins (12 from OmniPath). Network analysis using centrality algorithms and Random-Walk identified 168 core proteins. Drug simulation was applied to these core proteins to classify them as druggable or undruggable. Candidate drug-targets were prioritised using a combination of network and omics analyses, leading to in vitro validation experiments for 7 monotherapy candidates and the identification of previously unexplored combination therapy strategies.