logo apx for dark background



Spatial drug activity contexts, or DACs, capture how the various spatial organisations of the products of our genome cause drugs to have variable effects. To measure all DACs relevant for drug discovery, Apricot has developed a unique platform based on 17 years of pioneering R&D in the field of image-based systems biology. This high-throughput platform collects information from multiple spatial scales to train machine learning models that predict cellular responses to drugs with unprecedented accuracy. Apricot is generating a unique data resource in which multi-scale DACs are profiled for compounds covering a highly diverse chemical space with the aim to uncover generalizable predictive rules that connect molecular-scale predictions of compound-target interactions to higher-scale cellular responses and treatment outcomes. This will re-define drug discovery and personalized medicine approaches in the next decade.

High spatial resolution

The target of a drug can be located in multiple subcellular compartments where it interacts with different proteins. Subcellular DACs quantify how this affects the ability of a drug to inhibit or degrade its target. 

icon cell2

of many single cells

Cells are in different states that cause pathways to have different responses. 
Single-cell DACs reveal how the cellular state influences the downstream effects of drugs on various pathways.

icon cell4

within multicellular assemblies

The surroundings of cells vary a lot, and this causes cells to be in different states. 
Multicellular DACs define how drug responses are conditioned by the cellular microenvironment and tissue ecosystem. 

multiplexing GENOME products

Besides being spatially organized, cellular activities depend on the concerted activity of multiple components. This is captured by simultaneously multiplexing gene loci, transcripts, proteins, and protein states.

biovisioN across scales

To extract quantitative information from the resulting large bioimage datasets, custom-built computer vision methods are used that identify DACs at all scales relevant for drug activity, from single pixels to whole tissues. 

machine learning

The multi-scale measurements from DACs give unprecedented power to Apricot's machine learning algorithms to predict the responses of cells to drugs with high accuracy and generalizability.