Integrative discovery of treatments for high-risk neuroblastoma

Neuroblastoma (NB) cells exhibit a complex spectrum of pathway changes associated with oncogene activation, chromosome events, tumor micro-environment and super-enhancer states. So far, elucidating which pharmaceutical compounds could modulate the activation level of each known pathway in NB cells has not been feasible. We treated 2 patient-derived xenograft (PDX) NB cell lines with different chemical compounds, at 3 different doses (IC50, IC20, and IC10) and 2 time-points (6 and 24h). The whole-transcriptome of the treated cells was analyzed by a cost-effective DRUG-Seq method. Using factor analysis and supervised machine learning, the data was used to build a model that predicts the NB-specific outcome of 19,000+ drugs studied in other (non-NB) cell lines in several drug profiling databases (LINCS, CMap, DepMap). Overall design: Two patient-derived xenograft (PDX) NB cell lines treated with different chemical compounds, at 3 different doses (IC50, IC20, and IC10) and 2 time-points (6 and 24h). Please note that the D1, D2, and D3 (in the sample titles) correspond to IC10, IC20, and IC50, respectively.