Sequencing technologies allowing omics (genomics, transcriptomics, etc) analyses have evolved the past years enormously. These evolutions have fuelled the development of precision oncology approaches that use the molecular tumor landscape to guide clinical management. However, standard omics analysis of tumors is dependent on the availability of tumor tissue and is regularly performed on bulk tumor material ignoring the heterogeneous make-up of the tumor. Moreover, interpreting this enormous data flood remains the biggest bottleneck to bring precision oncology into clinical practice. With Non-Invasive Drug Identification for Tumors (acronym NIDIT) we will develop a novel analytical pipeline to tackle the current challenges: accurate driver gene identification, intra-tumor heterogeneity and biopsy accessibility. Optimized and validated using lung and solid pediatric tumors, the NIDIT pipeline will identify actionable targets and drugs for individual cancer patients to co-guide therapy stratification and decision-making. The pipeline will rely on RNA sequencing as this technique provides precise data for the expression of genes, alternative transcripts, fusion genes and sequence variants, moving beyond the scope of genomic data. From these data, patient-tumor-specific signature scores will be calculated for pathways inferred from a pan-cancer regulatory network and candidate target pathways or genes ranked. In a first step the method will be developed for application on bulk tumor samples. In a second phase, the method will be adapted for application on single-cell RNA sequencing from circulating tumor cells isolated from patient blood. These data are expected to illuminate vulnerable drug targets for the entire clonal population of a tumor. As circulating tumor cells are not always readily available in all cancer patients, we will also validate the pipeline on other blood fractions including circulating cell-free RNAs. The non-invasive nature of liquid biopsy reduces patient distress and allows sampling at any time from diagnosis to follow-up for complete care management. NIDIT pipeline predictions will be compared to standard methods for the identification of druggable driver genes using genomic data. Extensive validation will use large drug sensitivity datasets from cell lines and mouse models, and culminate in wetlab
drug prediction testing in patient-derived tumor xenografts. In conclusion, the NIDIT pipeline will provide robust non-invasive individualized drug prediction for any tumor type, providing proof-of-practice for a comprehensive computational strategy assisting precision oncology.