Emerging single-cell transcriptomic and highly-multiplexed imaging methodologies are advancing our basic understanding of tumor heterogeneity and its impact on patient outcome and therapeutic response. My previous work involved predicting disease progression and drug response using bulk (principally, expression) data. My interests lie in leveraging these new single-cell modalities to improve our ability to extract insight from the wealth of existing (and clinically annotated) bulk data. For example, we have recently completed a deconvolution DREAM challenge comparing methods that infer immune sub-populations from bulk expression data. Several participant methods used single-cell RNA-seq to identify markers that could subsequently be used to detect the corresponding population in bulk data.
