To mark Cellomatics’ 10 year anniversary, our CEO and Founder Dr Shailendra Singh shares his thoughts on the journey that Cellomatics has been on and what he has learnt over the last decade.
Over the past decade at Cellomatics Biosciences, we’ve been focused on a simple but important goal: improving how we model human biology in early drug discovery.
When we started, it was already clear that too many drugs were failing late in development. Ten years on, despite all the advances in science and technology, that problem still exists. A large part of it comes down to the same issue — early models often don’t reflect real human disease. 
Running a translational research lab for over a decade gives you a certain perspective. You begin to see patterns. You understand what genuinely helps move programmes forward — and what looks promising early on but doesn’t translate when it matters most.
Our work across organoids, primary human systems and translational biology has shaped how we think about this space. And while the science continues to evolve, some lessons have remained consistent.
Translational success starts with human biology
A lot of drug failures don’t start in the clinic — they start much earlier.
Traditional models like immortalised cell lines or simple 2D cultures have their place. They’re practical, scalable, and easy to standardise. But they are also inherently limited. They remove much of the complexity that defines real human biology.
Because of that, early data can often be misleading. Compounds appear effective in simplified systems, but when tested in more complex environments — or ultimately in patients — the effect is not replicated.
What we’ve consistently seen is that when you start with human-relevant systems, you get a very different level of insight. Using primary human cells, combining different cell types in co-culture systems, and applying disease-relevant stimulation allows you to better reflect how biology behaves in reality.
These models don’t just tell you if something works — they help you understand how it works and whether that mechanism is likely to translate.
This has a direct impact on decision-making. It allows teams to:
- Prioritise stronger candidates earlier
- Identify limitations before significant investment
- Reduce the risk of late-stage failure
In many ways, this is about shifting confidence earlier in the process. And the earlier that confidence comes, the more efficient the entire development pathway becomes.
Organoid drug discovery is changing the landscape
One of the most meaningful shifts we’ve seen over the past decade has been the rise of organoids and 3D systems.
Traditional 2D cultures provide useful information, but they flatten biology — literally and conceptually. Organoids, on the other hand, bring back structure, context and interaction.
They allow us to recreate aspects of real tissue in the lab. That includes:
- Spatial organisation of cells
- Diversity of cell populations
- Functional behaviour closer to what we see in vivo
This opens up new ways of studying disease. Instead of looking at isolated signals, we can start to observe how systems behave as a whole.
For example, in inflammatory or oncology settings, organoids allow us to study not just the response of a single cell type, but how different cells influence each other. That interaction is often where key disease mechanisms sit.
There’s no question that organoids are more demanding to work with. They require expertise, optimisation and careful handling. But the depth and quality of data they provide make that investment worthwhile.
For us, they represent more than a technological step forward — they represent a shift in mindset. A move away from simplification, towards biological relevance.
Multi-dimensional data is the key to success
Another major shift has been how we think about data.
Historically, drug discovery often focused on single endpoints — does a compound hit a target, does it inhibit a pathway, does it change a specific marker?
But biology is rarely that simple.
Over time, we’ve seen that relying on a single readout can lead to incomplete or even misleading conclusions. A compound may look effective in one dimension but fail when you consider the broader biological context.
That’s why a multi-dimensional approach to data has become essential.
By combining different layers of information — such as cytokine release, pathway activation, functional responses and morphological changes — we can build a more complete picture of how a compound behaves.
This is particularly important in complex diseases where multiple pathways are involved. Looking at one signal in isolation doesn’t capture that complexity.
Drug discovery is no longer about whether a compound hits the target, but whether it meaningfully influences human disease biology in a way that translates to patient outcomes.
Primary human cells introduce variability — and that’s valuable
One of the most common concerns when working with primary human systems is variability.
Different donors give different responses. Experimental outcomes are less uniform than in standardised cell lines.
At first, that can feel like a complication. But in reality, it’s one of the most valuable aspects of these systems.
Human biology is inherently variable. Patients respond differently to the same treatment. If our models don’t reflect that, we risk building false confidence.
By working with multiple donors, we gain a better understanding of:
- How consistent a drug’s effect is
- Where variability exists
- Which populations may respond more strongly
This kind of insight is difficult — if not impossible — to capture in simplified systems.
Rather than trying to eliminate variability, we’ve learned to work with it. To treat it as data, not noise.
This is where preclinical research starts to connect with precision medicine. Understanding variability early helps inform better clinical strategies later.
Scientific rigour cannot be compromised
As models become more complex, the need for rigour increases.
Human-relevant systems are powerful, but they are also more sensitive. Small variations in handling, timing or conditions can have a significant impact on results.
Over the years, we’ve learned that complexity alone does not guarantee better outcomes. It needs to be matched with strong scientific discipline.
That includes:
- Clearly defined and standardised protocols
- Consistent execution across experiments
- Robust quality control measures
- Careful interpretation of data
Without this foundation, even the most advanced models can produce inconsistent or difficult-to-interpret results.
True innovation is not just about building better models — it’s about ensuring those models deliver reliable, reproducible insights.
That’s what gives the data credibility.
The best work is always collaborative
Finally, one of the strongest lessons has been the importance of collaboration.
The most successful programmes we’ve supported have not been transactional. They’ve been partnerships.
When there is early alignment on the biology, clear communication around goals, and a shared willingness to adapt as data emerges, the quality of the work improves significantly.
In these collaborations:
- Study designs evolve as understanding deepens
- Questions are refined as new data becomes available
- Decisions are made with greater confidence
This creates a much more dynamic and productive process.
Preclinical research, at its best, is not just about generating data — it’s about contributing to scientific thinking and helping shape direction.
And that only happens through strong, open collaboration.
Final reflection: what 10 years has taught us
After a decade of working in translational human biology, one principle stands out clearly:
The closer your model is to real human biology, the more confidence you have in what comes next.
The field will continue to evolve. Organoids will become more refined, AI will help us interpret increasingly complex datasets, and regulatory frameworks are beginning to recognise the value of human-relevant systems.
But beyond the science, there is a broader shift taking place.
At Cellomatics, we are committed to supporting a more predictive and more ethical approach to drug discovery. This includes reducing reliance on animal models and improving the quality of early-stage research.
By generating reliable data earlier, we can reduce unnecessary experiments, focus on the most promising therapies, and improve development efficiency. This leads to better science and outcomes. The future of drug discovery lies in prioritising human biology from the outset, and we aim to support this shift by working with partners who share this vision.
Visit our news page to see the latest updates from the team at Cellomatics as we look towards the next 10 years as a company.
