Cellular life is complex. It is the results of a near-endlessly optimized dynamic molecular network where cells, their genomes, genes, proteins, metabolites, lipids, and the environment all continuously interact and influence each other in an intricate molecular dance. Complicating matters, we can typically only measure a small subset of the relevant players at the same time.
To decipher how cells work we therefore take a top-down point of view, focusing on combining large-scale functional data from image-based screens with large-scale molecular measurements, such as genome wide transcript abundance and quantitative lipidomics measurements. This integration ensures that we start from a functional point of view, asking what functional differences there are between cells or between individual people, and then apply unbiased OMICs approaches to find and validate those systems that best explain our observations.
The Human (C)hemotype Project
Major drug treatments are estimated to be effective in only 25 to 60% of patients, and millions of adverse drug reactions occur each year. Therefore, a better molecular understanding of what determines drug response is needed, and assays that can predict the drug response of an individual patient are of high clinical importance. Over the past years we have developed the technology to measure by automated microscopy the drug response of all individual peripheral blood mononuclear cells (PBMCs) from small single-donor blood samples over hundreds of drugs, which we call Pharmacoscopy. Pharmacoscopy now allows us to systematically measure drug response variability over individual humans, and statistically interrogate this variability to identify the molecular determinants underlying the drug sensitivity of a multilineage tissue at the single-cell level. Analyzing the variability in ex vivo drug response of PBMCs over a large cohort of healthy human donors using automated microscopy and single-cell image analysis will give the first insight into human-to-human variability of drug response at the single-cell level, and enable a classification of drugs based on the PBMC response profiles over healthy human donors.
As for any tissue, the cellular state and drug response of PBMCs is determined by a complex set of molecular and cellular factors, including environmental factors, genetic variations, gene expression profiles, metabolic activity, and cellular interactions. Statistical integration of the drug response profiles with basal gene expression and other molecular characterizations leads to concrete hypotheses as to the molecular determinants of the response, amenable to further in silico and experimental validation by biochemical and genetic means. We are thus working to advance the use of image-based screening of blood for basic research and personalized medicine, and reveal the structure and organization of drug response variability over individual humans and over hundreds of drugs, leading to a concretely improved molecular understanding of the determinants of drug sensitivity.