Understanding a drug’s mechanisms of action (MoA) is paramount to a drug discovery programme and represents one of the most critical challenges in pharmaceutical development. This fundamental knowledge - understanding how a therapeutic compound produces its effect through biochemical interactions with a target molecule - forms the cornerstone of a successful drug discovery programme. 

Even if a discovery team is certain about how a candidate molecule interacts with a specific target protein, how can they be sure that the interaction is creating the intended changes in target protein behaviour and subsequent biological manifestations? This intersection between mechanism of action and target biology represents a major blind-spot that is nearly impossible to assess, at least prior to entering the clinical phase of development.  

At MGB, we believe that spatial proteomics (captured at nanoscale precision) has the potential to transform how scientists interrogate this intersection between MoA and target biology. This represents a massive opportunity to move these critical insights into the earliest phases of discovery, during the critical window when programmes can still be directed towards clinical success and before major capital investments have been committed.

On this belief, Micrographia Bio, has developed the world’s first end-to-end platform for the quantification of changes in protein behaviour and MoA at scale, unlocking unprecedented insights to transform and power drug discovery programmes. 

Our platform technology combines three key capabilities:

1. Detection of individual protein instances at nanometer-scale precision
2. Detection of many different types of proteins in the same cell, enabling the direct, multiplexed visualisation of protein behaviour
3. AI-powered quantification of changes in protein behaviour and MoA, enabling the ability to distinguish MoAs across drugs, doses, or perturbations  

Together, these capabilities allow us to observe, measure, and quantify protein behaviour and drug MoA that were previously invisible using conventional imaging technologies.

Observing subtle changes at nanoscale resolution

Many drug effects occur in the cell at minute scales, or in tightly packed compartments like the nucleolus, in which commercial microscopy methods are unable to differentiate different molecules from one another. 

Our single molecule protein detection technology breaks through this barrier. By capturing multiplexed images at nanoscale resolution, we can examine the locations of multiple proteins simultaneously within the same cell, revealing the smallest changes in structural organisation, and protein behaviour / interactions in response to a drug treatment.

Quantifying changes in protein behaviour with AI

While visual inspection of images can reveal obvious qualitative differences, detection of subtle changes in this type of data is not humanly possible. Our AI-powered platform addresses this by embedding each spatial proteomic image into a high-dimensional space, capturing structural phenotypes and protein behaviour in a form that can be compared and analysed.

This gives us a quantitative endpoint for detecting cellular changes due to treatment. For example, we can assess how strongly a drug perturbs a given compartment, measure dose-dependent effects, or track protein dynamics over time.

Comparing MoAs directly

How much insight gets lost in discovery assays where samples are only compared to the control and reference samples? Understanding the difference from the negative control is the first step, but what scientists really want to know is, how are these samples different?  

Our approach allows for this assessment in every experiment context.  By comparing embedded phenotypes across samples, our model provides readouts of how drugs differentially modulate protein behaviour and cellular structure, and allows us to group, rank, and differentiate responses based on mechanistic similarity or divergence.

Our platform in action: differentiating MoAs in the nucleolus

The nucleolus is a small, dense, and highly structured organelle within the nucleus that is responsible for ribosome production. It consists of three main sub-compartments, each with specific functions and protein compositions. However, due to its complexity and small size, no existing methods are able to reliably visualise drug effects on nucleoli proteins and its organisation, creating a significant blind spot in drug studies targeting these organelles. 

We applied our platform to study the effects of two drugs known to act on the nucleolus: oxaliplatin and actinomycin D. Using conventional microscopy, these drugs are indistinguishable from each other, and from untreated cells.

‍

With our technology we are able to observe protein dynamics in the nucleolar compartments directly. At nanometre scale, clear differences between drug-treated and untreated cells emerge - and between the two drugs themselves.

‍

We then use our AI embeddings model to quantify these effects. Each protein marker is measured against the untreated state, producing an overall metric that captures the extent of the cellular changes seen in response to the drug.

‍

Finally, by comparing embeddings directly, we can confirm that oxaliplatin and actinomycin D induce distinct structural changes, quantitatively validating their different MoA.

A new lens for drug discovery

This case study illustrates the power of Micrographia Bio’s platform: to detect protein behaviour and subtle, spatially complex changes; to quantify them in a reproducible way; and to distinguish drug behaviours at scale. By transforming spatial images into structured data, we enable new kinds of insight into how drugs work, and open new opportunities for safer, more targeted therapies.