Bonus Episode: Statisticians vs Biologists

We’re super excited to kick off a brand new season with a super special guest.This time, Markus and Phil sit down with Aziza Yormirzaeva to talk DOE, collaboration, failure, and AI.What happens when your experiment “fails”,  but actually tells you something more important?Plus: why scientists often think DOE means more experiments, and whether AI is coming for Aziza’s job 👀

“We compress DNA from oil-tanker scale to 96 wells - and still capture everything that matters.”🎙️ The latest episode of The Next Experiment is out!Markus Gershater and Phil Kay sit down with Claes Gustafsson to talk about the power of design of experiments (DOE) in biology, how to get the most out of your data, and why smart experimentation beats brute force every time.Claes shares how DOE lets you compress enormous experimental spaces, optimize proteins, and even improve enzymes used in millions of cheeses worldwide. He dives into codon optimization, iterative rounds of testing, and why a good assay is always king.Some standout lines:🔹 “If your assay sucks, it doesn’t matter how much DOE you put on it - it will still suck.”🔹 “Biologists love collecting data, but more data isn’t better. The question is: how do you get the most bang for your data buck?”🔹 “You don’t need to understand every mechanism. You just need to know which experiments will reliably get the outcome you care about.”If you’re in biotech, protein engineering, or any field where structured experimentation matters, Claes’ insights are gold.

“Most assays don’t fail because they’re wrong… they fail because they’re fragile.”This time, Markus Gershater and Phil Kay sat down with Rob Howes (Head of Biology & DMPK at Charles River) to unpack what REALLY happens when you scale biology - and why most teams underestimate how quickly things fall apart.Rob’s been working with multi-dimensional experimentation since the early 2000s (back when most of us were still painfully tweaking one variable at a time). And the contrast is pretty stark:Some standout lines:🔹 “We used to optimise for the highest signal. Now we optimise for something that won’t break.”🔹 “If your data isn’t high quality, your AI model won’t be either. Garbage in, just faster garbage out.”🔹 “Why run every experiment, if you can predict most of them?”

“Some people don’t want experiments to be repeatable, they want them to feel like magic in their hands.”🎙️The latest episode of “The Next Experiment” is here!This time, Markus and Phil sat down with Dr. Linden Schrecker, co-founder of SOLVE Chemistry, to explore the future of experimentation, automation, and why better data changes everything.From failed reactions caused by cold Oxford labs to disproving decades of accepted chemistry with a single well-designed experiment, Linden shares a fascinating perspective on how scientists should think about experimentation in the age of AI.The conversation dives into reproducibility, metadata, machine learning, and the cultural incentives that often prevent researchers from collecting truly meaningful data.Some standout lines:🔹 “You can collect 50 experiments in the same time as one, if you design the system properly.”🔹 “We’re not incentivising scientists to collect meaningful amounts of data. We reward novelty, not repeatability.”🔹 “We should be collecting data for a future we may never personally benefit from.”If you’re working in chemistry, biotech, automation, AI, or process development, this episode is packed with insights on how smarter experimentation can unlock entirely new ways of understanding complex systems.

“Drug-resistant bacteria played the ultimate uno reverse card on us.”🎙️The latest episode of “The Next Experiment” is here!This time, Markus and Phil sat down with Adam Winnifrith - who shared his unique perspective on experimentation, scaling, and why understanding your system is always step one.Adam dives deep into the realities of doing science at scale, the pitfalls of jumping straight to automation, and how small, smart experiments can save months of wasted effort.Some standout lines:🔹 “You can automate all you want, but if you don’t know what’s actually happening in your system, you’re just speeding up failure.” 🔹 “DOE isn’t just a tool; it’s a mindset shift. It forces you to question assumptions before spending a month chasing the wrong lead.” 🔹 “The real magic happens when you combine curiosity with disciplined experimentation.”If you’re in biotech, research, or any field where experimentation is key, Adam’s insights are not to be missed.

"I wouldn’t describe DOE as easy — I’d describe it as extremely worthwhile.”🎙️The latest episode of “The Next Experiment” is here!This time, Markus and Phil sat down with Alex Rimmer, who brings a clear and practical perspective on DOE in biology – and some truly memorable lines about why experimentation so often fails to scale.Alex talks about discovering DOE out of necessity, not theory. When nothing else worked, DOE became the only way to make progress. A few great quotes:🔹 “Biology is basically the study of interactions that we largely don’t study interactions for.” 🔹 “I had a need to make a media and absolutely no idea how to get that to work via traditional non-DOE methods.” 🔹 “Automation on its own won’t be enough; it’s what you apply with that automation that really matters.”If you care about scalable experimentation, media development, or the practical reality of doing modern biology, this episode is full of insight.

This time, Markus and Phil sat down with Owen Jonathan, who brings a refreshingly open and grounded perspective on learning – and teaching – DOE in real biological systems.What makes Owen’s story so compelling is how honestly he talks about coming into DOE with an open mind, realising what he hadn’t been taught, and then discovering how powerful the methodology becomes when biology is too complex for intuition alone.A few standout quotes:🔹 “I didn’t realize until later that I’d never actually been taught DOE.” 🔹 "Progress isn’t about doing more — it’s about asking the right questions first." 🔹 “Biology needs abstraction or it becomes impossible to model.”If you’re interested in practical experimentation, ML-enabled workflows, or how real scientists bring DOE into messy biology, this one’s packed with insight.

If efficiency is the name of the game in process development, then Design of Experiments (DOE) is the playbook you didn’t know you needed. But how do you actually apply it in the messy, variable world of viral vector production? That’s exactly what we dig into in this episode. Keara Sutherland takes us through her journey from simple drug pairing experiments to complex, high-dimensional DOE campaigns that optimize biological processes in ways that feel almost magical – but are grounded in solid strategy.Conversation highlights:05:24 - Keara’s first exposure to DOE and transitioning from drug synergy studies to complex process optimization12:17 - The addictive efficiency of DOE: maximizing data points while minimizing time17:26 - DOE as a tool for biological flexibility: managing multiple outputs and future-proofing processes31:31 - Real-world challenges: balancing experimental ideal with practical constraints, including automation and scalability

Making multidimensional experiments a reality doesn’t just sit with one discipline.  Most commonly, it involves biologists and statisticians. So, what’s the key to open communication, and working together to do radically better biology? That’s what we explore, in this season’s bonus episode, filmed at Synthace’s Beer, Bytes and Biology annual event.  Conversation highlights 00:00 - Introduction 00:49 - Condescending attitudes causing statisticians and biologists to butt heads  02:49 - The key to successful relationships between the 2 disciplines 04:09 - Markus’ experiences showing how he and his team bridged the communication gap  08:39 - Embodying the “George Box” attitude for successful collaboration 

If data is the new oil in an AI centric world, then amassing ever-larger multidimensional datasets can only be a good thing.  But how can we use these datasets to gain deeper insight into biology? This is the question we try to answer in season 1’s final episode. Between us, we bring views of the AI optimist and skeptic—with starry-eyed visions and sobering realities—to the table before reaching a conclusion. Conversation highlights 00:00 - Introduction 01:18 - A vision for multidimensional datasets fueling better experiments 03:37 - AI and myth-busting: “It’s not magic”   09:15 - Risks of combining datasets for making “pizza” and “ice cream” 10:11 - Ideals and potential low-hanging fruit for AI in biology 13:09 - Thinking about AI as collective memory18:16 - Combining human and artificial intelligence is key