I’ve seen people go through a cancer diagnosis and honestly it changed how I think about medicine entirely.

The oncologist ordered a genomic panel. Two weeks later, the results came back showing a specific mutation that ruled out the standard first-line treatment, the one most patients with that diagnosis receive and pointed toward a targeted therapy that would have been considered experimental a decade earlier. The difference between those two paths, in terms of outcome probability, was not marginal. It was enormous.

What struck me was not the technology. It was the “implication” that the “standard” treatment, the one that had been refined over decades of clinical trials on large populations, was simply the wrong treatment for this particular person. Not because it was bad medicine. Because it was designed for an average that this person did not fit.

That is the problem personalised medicine is solving. And the reason it is gaining attention so rapidly is that the tools to solve it have finally arrived at scale.

Modern medicine was built on population thinking. You run a trial on thousands of patients, measure the average response, and design a protocol around that average. It is logical, rigorous, and for a large portion of patients it is reasonably effective. 

But medicine has always known that the average patient is a statistical abstraction. The actual patient sitting in the clinic is a specific human being with a specific genome, a specific microbiome, a specific history of exposures and stresses and nutritional patterns that shape how their body will respond to any given intervention.

Ignoring that specificity was never a choice born of indifference. It was a constraint of capability. Until recently, sequencing a genome cost tens of thousands of dollars and took weeks. Running a comprehensive biomarker panel was the preserve of the best-funded research hospitals. The data infrastructure to make sense of individual variation at scale simply did not exist.

All three of those constraints have collapsed in the last five years. Whole genome sequencing now costs under $300. Liquid biopsies that detect disease-specific biomarkers from a blood draw have moved into clinical practice. And AI platforms can now cross-reference an individual’s biological profile against millions of data points to suggest treatment pathways tailored to that specific person.

The Trial Design 

Clinical trial design is the most conservative, most carefully regulated part of medicine as it moves slowly because the stakes are too high to move fast. The fact that biomarker-stratified and adaptive trials have gone from 21% of new registrations in 2018 to more than 58% in 2026 is not a reflection of fashion. It is the scientific establishment acknowledging, through its most fundamental tool, that population-averaged evidence is no longer sufficient.

Adaptive trials are designed to learn from patients as the trial progresses adjusting dosages, stratifying subgroups, dropping arms that are not working and reinforcing ones that are. Biomarker-stratified trials recruit patients based on specific biological characteristics rather than just diagnosis. Both approaches treat the individual as the unit of analysis rather than the population.

When the methodology of science changes, the medicine that follows changes with it.

The Concept of Personalisation

It is not unique to human medicine. It is a principle that runs through all biological systems and it is one that shapes our thinking at TerraPHA deeply.

We work at the upstream end of health be it in aquaculture, in animal nutrition or in soil biology. And the shift we are seeing in those domains mirrors, almost exactly, what is happening in human medicine. The old approach to aquaculture health was prophylactic; dose the pond with antibiotics, control the environment by chemical suppression, treat every population the same way. It was the equivalent of one-size-fits-all medicine. 

It worked until it stopped working, and the signs that it has stopped working were antimicrobial resistance, biosecurity failures, declining yields which are now impossible to ignore.

What is replacing it is precision and understanding that the specific microbial composition of a given pond, the specific nutritional deficits of a given batch of feed, the specific stressors present in a particular production environment. Our PHA-based bio-inputs are not designed as blanket solutions. They are designed to work with the biology that is already present to restore balance and resilience at the microbial level, rather than override it chemically.

What This Means Beyond the Clinic

I build homes as well as biology, and even in that world the personalisation signal is getting louder. The buyer in 2026 does not want what the buyer in 2010 wanted, a standard 2BHK designed by a formula that has not changed since liberalisation. They want spaces that flex to their life, a home office that actually works, a layout that accommodates ageing parents and growing children, a building that is managed responsively rather than according to a schedule set in 1998.

The desire to be treated as an individual rather than a category is not a healthcare phenomenon. It is a civilisational one. Personalised medicine is simply its most scientifically consequential expression, the domain where the stakes of getting it right are measured in years of life rather than years of a lease.

That is why it is gaining attention. Because it is long overdue, and because the tools to deliver it have finally caught up with the idea.