Can every drug target be imaged?

Behind the Paper: Repurposing molecular imaging to map drug targets in vivo

Martin Walter Jul 21, 2026

Scientific projects rarely begin with grand visions. More often, they begin with a simple question that refuses to go away.

For us, that question emerged from everyday clinical practice in molecular imaging. Modern medicine increasingly relies on targeted therapies. We prescribe drugs that inhibit specific receptors, enzymes, signaling pathways, and molecular mechanisms. Yet in many cases, we have no direct way of visualizing these targets in living patients. We can often see the tumour, the inflammation, or the disease process, but not necessarily the molecular target that we are trying to treat.

This apparent disconnect between therapy and imaging became the starting point of a research journey that ultimately led to our recent paper, Repurposing molecular imaging to map drug targets in vivo.

The story actually began several years earlier.

Our first question was straightforward: How much of human biology can be visualized using molecular imaging?

At the time, molecular imaging felt fragmented. Individual tracers had been developed for individual targets, diseases, and clinical questions. What was missing was a broader perspective. We therefore decided to systematically map molecular imaging targets across the human genome. The result was what we called The Imageable Genome.

To our surprise, the imageable landscape was far larger and more diverse than we had anticipated. Instead of a small collection of isolated targets, we found a complex network spanning hundreds of genes, pathways, and disease processes. For the first time, we could begin to view molecular imaging as a system rather than a collection of individual tracers.

But one question immediately followed another.

Many imageable targets are also therapeutic targets. If imaging and therapy increasingly converge, how large is this overlap? Which targets can both be visualized and treated?

This became the basis of our second project, The Theranostic Genome. In that study, we systematically mapped the intersection between imaging and therapy. Again, the picture that emerged was larger than expected. The boundaries between diagnosis and treatment appeared increasingly blurred, and molecular imaging seemed positioned to play a much broader role in precision medicine than traditionally assumed.

Yet an important limitation remained.

Most approved drugs target proteins that cannot easily be imaged directly.

Some targets are intracellular. Others lack suitable tracers. Many have simply never been considered imaging targets. If direct imaging were the only option, large portions of modern pharmacology would remain invisible.

At first, this appeared to be a dead end.

Then transcriptomic data offered a different perspective.

Genes rarely function in isolation. Instead, they operate within networks of co-expression and coordinated biological programs. This led us to ask a deceptively simple question:

What if a drug target does not need to be imaged directly?

What if it could be visualized indirectly through a biologically linked imaging target?

Once we began exploring this possibility, the project rapidly expanded. What started as a simple question gradually evolved into a much broader effort to connect molecular imaging with modern pharmacology.

At some point during the project, we realized that we had stopped thinking about individual tracers altogether.

Instead, we were thinking about relationships.

Relationships between diseases and drugs.

Relationships between drug targets and imaging targets.

Relationships between molecular pathways and imaging agents.

The project had evolved from a study of tracers into a study of connectivity.

One of the most rewarding moments came when we assembled the final resource. For the first time, we could follow a chain from a disease, to an approved drug, to its molecular target, to a directly or indirectly imageable target, and finally to one or more available radiotracers.

What emerged was not simply a database.

It was a navigable map connecting the currently drugged portion of the druggable genome to the imageable genome.

Of course, many challenges remain. Co-expression does not automatically imply biological equivalence. Imaging hypotheses generated computationally require experimental validation. Not every proposed imaging strategy will prove clinically useful.

Nevertheless, the framework suggests that molecular imaging may be able to reach far beyond the relatively limited set of targets that can currently be visualized directly.

Perhaps the most exciting aspect of the project is what comes next.

For decades, molecular imaging has largely developed tracer by tracer, target by target, and disease by disease. Our work hints at a different future in which diseases, drugs, targets, imaging agents, and treatment outcomes can all be connected within a single framework.

Looking back, the journey was shaped not by a single breakthrough, but by a series of increasingly ambitious questions. Each answer revealed an even larger landscape to explore.