The battle lines are drawn and the ethanol industry’s equivalent of the Met Gala is about to begin. When June rolls around, the titans of the enzyme and yeast industries roll out their latest offerings. The strain engineers unleash their latest offerings on the runways, and we all sit back and marvel at the audacity, the vision, the skill. And, since it’s all about Performance, it’s Performance Art of the highest order
All companies are furiously innovating — consider, for example, the Lallemand advanced yeast that started the Yeast Wars. Leaf’s been at it hammer and tongs, too. Iron sharpeneth iron. Innova Element from Novonesis answered a clear need for better performance and reliability, and its Fiberex F2.5 continues to advance its fiber-degrading technology. What next will we see from the Meisterbrewers of the bioeconomy?
For sure, we are going to see amazing work in rate, titer, yield, target. Those are traditional metrics that matter for a reason. And when it comes to strain and process engineering, we’re probably in safe hands if we are looking for a couple points of improvement every year. Or, a novel product we can make from dextrose. Or, something that will last longer.
But true, step change improvement, is it possible? Today, let’s explore the idea that some outside-the-box-thinking might provide some outside-the-box performance. And, we’re going to start with the idea that underlying performance, underlying robustness, is persistence. The organism that will not die, that finds ways to succeed, that becomes a platform, a class, an enduring hit. That’s what we all want, isn’t it? Not success of the hour, but success for all time.
Where might we first look? I said “out of the box,” but this isn’t unfamiliar territory. It’s something you see and use every day… it’s the ultimate industrial processing system. It’s been around 15 billion years and still runs great. Just wave your hand through the air, that’s space, the period in which you hand moved, that’s time, and today we’re going to look at what we can learn from the world’s oldest and most successful processing system, the source of space and time, the very source of strains, to improve strain engineering.
There aren’t too many manuals on spacetime as an industrial processing and strain design system, so we may be looking at familiar things in unfamiliar ways, and unfamiliar things in familiar ways. But I’ll try to keep it friendly, not too math-y. For now, just remember, space and time haven’t been around forever, they’ve been around 15 billion years, they are a solution to a problem; they are so good at solving this problem, we don’t even know it’s there. We’re going to look just a little at where they came from, to apply their lessons to synthetic biology.
Let’s start here, on well-trodden ground, with a picture of sacchoromyces cerevisiae, the beautiful and ancient workhorse of enzymes, friend of glucose, pal of ethanol, the fellow who put the rise in your bread.
Our little yeast friends make ethanol from sugar, very well. They have challenges. The mother of their challenges? It’s heat. Our organisms are extremophiles — but ultimately all organisms falter, break down, die when the heat’s turned up. It is not a question of whether, but when.
So, we already have a clue that many of the answers to our questions lie in the realm of thermodynamics, which, like genetics, is generally speaking in mathematical terms, part of the world of number theory and analysis. Yet, we might also guess that when we design for speed, shape comes into the equation in synbio as well as inorganic catalysts. That’s the world of topology.
It is the study of shape and form. It is not the world of bases and acids but the boundaries, surfaces, depths, folds. It is at the heart of catalysts. To illustrate, consider this illustration of yeast buds.
Then consider this.
Pretty good model of the living system above, right? We see the budding, the dividing. Only, this last illustration are not models of yeast division and budding. These are stages of a Ricci flow on a two-dimensional manifold, a partial differential equation for a Riemannian metric. In other words, deeply part of the world of Euclidean geometry and topology, and not usually part of the yeast discussion.
Are these geometric and biological anomalies? Not at all. You can see Ricci flows like this in raindrop formation.
We see them in binary star formation, too.
I promised we wouldn’t spend much time wandering the cosmos, so we’ll pause there on binary stars and raindrops. Let’s retreat back to the safety of emergence, budding, catalysts and membranes. Let’s start with the most common atom there is, and foundational to organic chemistry, our friend hydrogen. Looking real close, you’ll see this.
Now, let me model that for you, to clean up the image a little, using a Ricci flow. Here’s one I’ve diagrammed for you, some 0s and 1s in a scenario, and I’ve added a permeable membrane that our object is able to cross.
Overall, pretty familiar, you have a center, a darker shell, and a light almost diffused outer shell. Not with the usual sharp boundaries of the human-sized world, but with some of the fuzziness of the atto-scale. Pretty close. Now, two more images I’ll share with you, then let’s reflect, OK? First, here’s the double helix we know well from the study of DNA.
Now, here’s something closely related to a double helix, but not exactly, exactly the same. It’s a helicoid.
The interesting thing about the helicoid that I asked my model to diagram the most thermodynamically efficient shape it could come up with for processing information. Now, look at the helicoid again, in your mind like look at the boundary edges, eliminate the stuff in between. What you get is the double helix.
And if you begin to consider that what biology is doing is finding an efficient way to process information at a thermodynamic level — and that what life is somehow fundamentally about is processing something important and precious — well, we’re traveling in a well aligned way, you and I.
Now, let’s drop back for a second to my Ricci Flow model diagram, here it is again.
You’ll notice that there’s quite a bit of gobbledygook about symbolic this and thatm that seems to have absolutely nothing to do with modeling hydrogen. Well, yes and no. For sure, if you’ve surmised that we are not looking at the attoscale, or life or membranes at all, well done you.
What I’ve dialed up for you reaches back to our most successful industrial processing system, the 3D observable universe. Above, this is a modeled diagram of the Big Bang, what you are seeing is all the energy that ever was or will be, emerging from a gate and into our known world. It’s not nanoscale, it’s everything-scale. Yet, decidedly familiar, looks a lot like our friend hydrogen, doesn’t it, and not unlike raindrops or yeast buds? Turns out, if you think about the Big Bang as an emergence event instead of an origin event, it looks an awful lot like synthetic biology.
You’d be surprised what a membrane can do, in this case an L-Gate, as a processing environment. Well, if you are in the world of fermentation, you might not be — membranes are part of the everyday conversation about chasing efficiency.
I recommend this particular one, the L-Gate, for study. It operated for less than a nanosecond, and produced all the complexity we have around us. There’s a big bunch of inputs, and passing through this gate, you get all the energy and all the instructions that will ultimately result in particles, molecules, DNA, people, animals, the whole shebang. This is not literally, like God creating the animals and the Garden of Eden, and such. If you’ve assembled complex items from IKEA, you can relate. And it’s not entirely unlike the instruction set known as DNA.
Here today we’re looking towards strain engineering, so we’re going to skip past the L-Gate and get to the implications and the conclusions, what you can do with this framework. However, for the curious, let me explain that an L-Gate is a zero-dimension singularity, permeable at a zero kelvin event. There’s a bucket of math if you want to dive in, happy to share: if you like Euler’s Identity, you’ll find it pretty easy to grasp. But our column today is about strain engineering and lessons that topology can give us.
Here are some high-level takeaways to spare you a math deep-dive. First, let’s look at topology and make the connection from the biggest structures to the smallest.
| Shape/Pattern | Biological Role | Emergence Analog |
| Sphere / bud | Yeast cell division, raindrop | Stable loop → spiralized matter |
| Rod / teardrop | Star formation, fluidic split-off | Transition wave → extended structure |
| Helix / helicoid | DNA, heat transfer | Minimal surface encoding → memory/code |
| Fractal/filamentous | Endoplasmic reticulum, fungal growth | Recursive persistence → computation/storage |
| Ricci surface flow | Heat diffusion | Symbolic stress curvature redistribution |
Now, let’s look at relating some conceptual terms between the sky and the inside of our fermenter.
| Classical Concept | Chemical Equivalent | Process Engineering Analog |
| Time (t) | Rate | Reaction rate / processing speed |
| Space (x) | Titer | Volume, concentration (extent of medium) |
| Matter (m) | Yield | Output product, persistence of form |
| Energy change (ΔKE) | Exported entropy | Work extracted from symbolic transformation |
| Information structure | Compression efficiency | Degree of meaning retained |
So, that’s why it’s very useful to think of the cosmos, really, as a reactor system converting symbolic potential into persistent structure. Slightly different feedstocks and products, but the same system. And, you’ve heard about expanding space, think dilution. Just as yeast converting sugar into ethanol reaches a point where it poisons itself, turns out, the cosmos needs dilution too —space, liquid, processing slack.
Now imagine the universe as a bioreactor for symbolic compression, converting quantum potential into structured persistence. As symbolic residues (information) build up, conversion becomes inefficient. To continue, the system dilutes itself—expands the processing field. This flips the cosmological model in a way that makes lots of phenomenon easier to map and explain. For example, cosmic voids. I’ve run some very good simulations with cosmic void data to confirm that this framework makes confirmable predictions, Happily, it does.
Practical Implications for Strain Engineers
So what does this mean in the here and now—for the engineers tuning valves, measuring titers, and chasing robustness?
First, it means you’re already working with cosmological tools. Rate, titer, and yield—these aren’t just fermentation KPIs. They’re local expressions of a much deeper thermodynamic behavior: the universe’s own method for exporting entropy and compressing meaning into persistence.
Robustness, as we commonly understand it, is microbial survival under stress. But persistence is a richer, more dynamic target. It includes:
- How long an organism continues to operate near peak TRY.
- How well it resists symbolic drift—unintended pathway mutations, trait collapses, and data loss.
- How predictably it performs when scaled across reactors, sites, or seasons.
In this framing, a high-persistence strain isn’t just strong—it’s directional. It doesn’t just resist stress; it uses stress to channel more energy into useful structure. It’s efficient not just in output, but in thermodynamic storytelling—retaining information through stress, compression, and recovery.
A New Aiming Point: The Persistence Quotient (PQ)
To that end, let’s offer a metric. If we want to measure persistence, we need to combine performance and shape. Here’s a first pass:
This gives you a normalized measure of how much structure (TRY) is preserved over time despite system stress, relative to how much symbolic disorder is introduced. The lower the drift, the higher the PQ. The longer a strain stays near the top of its curve—without requiring human correction—the better.
In plain language: PQ measures how hard the strain works, how long it works that hard, and how much entropy it sheds doing it.
The Takeaway
For synthetic biology and strain design, this is a chance to pivot from optimizing output, to engineering shape. From minimizing failures, to maximizing resilience. From batch-to-batch improvement, to designing for persistence across time and complexity.
What can’t be measured can’t be improved. But now, perhaps, we can measure persistence—and build toward it, from the nanoscale to the cosmos.
About the Author
