Signaling underlines nearly every cellular event. Individual cells, even if genetically identical, respond to perturbation in different ways. This underscores the relevance of cellular heterogeneity, in particular in how cells respond to drugs. This is of high relevance since the fact that a subset of cells do not respond (or only weakly) to drugs can render this drug an ineffective treatment. In spite of its relevance to many diseases, comprehensive studies on the heterogeneous signaling in single cells are still lacking.

We have generated the, to our knowledge, currently largest single cell signaling dataset on a panel of 67 well-characterized breast cancer cell lines by mass cytometry (3’015 conditions, ~80 mio single cells, 38 markers; Bandura et al. 2009; Bendall et al., 2011; Bodenmiller et al., 2012; Lun et al., 2017; Lun et al., 2019). These cell lines are, among others, also characterized at the genomic, transcriptomic, and proteomic level (Marcotte et al., 2016).

We ask the community to use these measurements to predict the signaling responses of the cell lines to drug treatment at the single-cell level. Firstly, we ask to predict certain markers in specific single cells. Secondly, we ask to predict the time-dependent response of single cells upon treatment with drugs. Finally, we aim to predict the time-dependent response of cell lines for which only static, unperturbed data is given. The methods developed to answer these questions will allow us to better understand the determinants that control single-cell signaling, the heterogeneity in drug response of cancer cells, and to push the limits of signaling modeling.

Related publication: