Abstract: In this talk I will summarize our recent and ongoing work analysing large pharmaco-genomic screenings in cancer, that provide rich information about alterations in tumours that confer drug sensitivity or resistance. Integration of this data with prior knowledge on signaling pathways and transcription factors provides biomarkers and offer hypotheses for novel combination therapies. Multi-task learning combining these features and drugs’ information leads to further insights into drug efficacy. I will also describe the insights gained when addressing this problem in a crowdsourced manner, by posing a challenge to the scientific community. This was performed in the context of the DREAM Challenges, and I will present this initiative and its infrastructure by means of containerized model submissions, that allows us to use sensitive data for the challenges. An important message from the challenge and our own results is that that prediction of drug efficacy is far from accurate, implying important limitations for personalised medicine. An arguably under-appreciated aspect that deserves further attention is the dynamics of signaling networks and how they response to perturbations such as drug treatment. I will show how cell-specific logic models, trained with measurements upon perturbations, can provides new biomarkers and treatment opportunities not noticeable by static molecular characterisation. Finally, I will show how, using novel microfluidics-based technologies, this approach can also be applied directly to tumor biopsies.