If you could run an experiment in your computer instead of the lab—and it actually gave you answers worth acting on—would you try it?

Biologics formulation is often described as a high-stakes puzzle. Every recombinant drug is a chemical balancing act: choosing the right excipients, predicting stability, and sidestepping months of trial-and-error. But what if you could speed things up with a virtual test drive?

In this episode, David Brühlmann sits down with Giuseppe Licari from Merck Healthcare, whose expertise is quietly reshaping how proteins reach the clinic.

Giuseppe Licari brings a hands-on perspective to computational formulation development. With a track record in applying molecular dynamics simulations to real-world drug development, he’s not just theorizing about the future—he’s showing what’s possible now.

Key Topics Discussed

• Biologic formulation is complex, with many components and conditions to navigate.
• Molecular dynamics helps visualize protein behavior and protein–excipient interactions.
• Computational tools can’t simulate long-term stability directly, requiring clever bridges to real timelines.
• AI and machine learning offer promise but still depend on large, high-quality datasets.
• Future workflows may use in silico tools to choose formulations first, with experiments mainly for validation.
• Smaller companies can access computational methods through SaaS platforms, balancing cost and expertise needs.
• Growing GPU power enables simulations of larger, more realistic systems.
• Integrating in silico approaches and staying open to new technologies is key for future formulation success.

Episode Highlights

• How molecular dynamics simulations are used to understand protein behavior and excipient interactions during formulation [03:03]
• The limitations of computational methods for real-time stability studies and how they can still inform long-term stability predictions [05:08]
• Where AI and machine learning are currently being applied in formulation development, and their future promise as datasets and computational power improve [06:36]
• Strategies for smaller biotech firms to leverage computational modeling, even with limited budgets or specialist staff [09:55]
• The importance of staying curious, open-minded, and proactive about adopting new computational approaches [12:08]

In Their Words

We’ve seen that in silico methods have been around for many years, and they are now a standard tool to support our work across several steps of drug discovery and development. I think they’re here to stay for the years to come.

So my message is: don’t be afraid to use them, explore them, and be curious. If these methods are helpful, why not embrace them?

Episode Transcript: From Developability to Formulation: How In Silico Methods Predict Stability Issues Before the Lab - Part 2

David Brühlmann [00:00:36]:
Welcome back to Part Two with Giuseppe Licari from Merck Healthcare, where we’re tackling the toughest formulation challenges in biologics development. Check here part one of our conversation.

How do you predict aggregation before manufacturing? What can simulations tell us about excipient-protein interactions? And when is it time to stop computing and start experimenting in the lab?

Giuseppe shares practical workflows, real success stories, and honest limitations of computational approaches. Plus, he’ll give one actionable step you can start using tomorrow.

David Brühlmann [00:01:14]:
Let’s jump back in. So we’ve done our homework: we’ve shown that our molecule is developable, we’ve assessed formulatability, and now we’re developing the proper formulation for the recombinant drug.

What are the specific in silico approaches you’re using? It feels to me like a very difficult puzzle — so many chemical components, different concentrations, and combinations. How do you find the needle in the haystack?

Giuseppe Licari [00:03:03]:
First of all, it’s important to remember that for formulation development, you need to look at how the protein behaves in its environment over time.

You can’t base your assessment on a static picture of the antibody — you need to watch the “movie.” In our field, this is generally done using molecular dynamics, a technique in computational chemistry that allows you to see how molecules move. You can literally see the protein dancing, if you want to imagine it like that, and observe how its conformation changes over time.

When you add excipients or buffers, you can see how those elements interact with the protein. From these interactions, you can extract conclusions about how the excipients might affect protein stability or alter its properties.

This is a critical point: looking at the protein “in motion.” In many ways, it’s like performing an experiment — but computationally. You simulate what happens in the lab: taking your drug substance, putting it in a specific environment, and observing its behavior over time.

Of course, it’s not exactly the same as the lab, but it’s a semi-realistic representation of reality. And it can still provide valuable, actionable insights that help guide your experimental work.

David Brühlmann [00:04:49]:
And how do you go about real stability studies? This is what takes time — you can’t compress it. Well, obviously, you can use stress conditions, but it still takes time to confirm outcomes. How do you combine in silico methods with real stability studies?

Giuseppe Licari [00:05:08]:
Of course. Real-time stability studies, like those used to assign shelf life, can’t be directly computed or simulated using in silico methods. One limitation is that simulations can only cover short periods of time — you can’t simulate six months of stability. So, that’s out of the scope of computational methods.

However, long-term protein stability is closely tied to the intrinsic properties of the molecule. That’s what we aim to study: which molecular properties correlate with long-term stability. Once you understand these connections, you can tweak formulations to adjust the protein’s behavior and get an estimate of long-term stability, even if the simulation only covers a short time. This is the strategy we typically follow.

David Brühlmann [00:06:11]:
Looking ahead, our industry is evolving rapidly with all kinds of technologies — AI, machine learning, and so on. Where do you see formulation development going in the next few years? How will we develop formulations for recombinant proteins?

Giuseppe Licari [00:06:36]:
In the AI space — with machine learning — there are several efforts to predict protein formulation behavior. You need a significant amount of data to predict outcomes, like aggregate levels or low molecular weight species. One current limitation is that we don’t yet have enough data to build models that are robust across many proteins and systems.

That said, AI is already helping in discovery. Generative AI can design proteins with fewer chemical liabilities and improved developability. Improvements early in discovery will have significant downstream effects, including on formulation development and other steps. Optimizing proteins from the start can make the entire process faster and more efficient.

David Brühlmann [00:08:06]:
Yes, and with more advanced generative AI models and more powerful computational techniques, you might even be able to select the optimal sequence and predict its ideal formulation — if I’m thinking futuristically.

Giuseppe Licari [00:08:27]:
Absolutely. I’m looking forward to when in silico methods can predict the optimal formulation and experiments are only needed to confirm the predictions. Right now, we still need some screening and lab tests, but in a few years, we might be able to reduce lab work significantly. This will reduce timelines, lower costs, and allow us to develop more molecules for patients more efficiently.

David Brühlmann [00:09:10]:
And to make our conversation very actionable. I would like now to look into how, for instance, someone who is working in a smaller company could apply that. Because the challenge is always, especially as we look in the future. It’s exciting, we have amazing new technologies. You could do a lot of things. And I think in a larger company — that’s also my experience — you’re pretty fortunate to have a lot of resources. But when you’re working in a startup or a small-to-mid-sized company, you have more limited resources. So what would be your advice to still leverage at least some of the potential of these in silico approaches, even with a smaller budget or without all these specialists in-house?

Giuseppe Licari [00:09:55]:
Yeah, sure. That could of course be a problem for small companies or startups. Maybe the solution would be to do a small feasibility study with an external provider. Nowadays, there are more and more companies providing software-as-a-service, for example, so you can test these approaches through a third party and see if they provide additional information or valuable outcomes for your project.

That’s something achievable even for smaller companies, because in silico methods are generally not very expensive computationally. You don’t need to invest too much to test a few things. So my advice would be to test with external companies and see if it works. Of course, the best solution is to hire a computational scientist internally to really build internal knowledge. It’s always better to have someone in-house, but we need to make compromises all the time.

David Brühlmann [00:11:02]:
Definitely. Absolutely. Before we wrap up, Giuseppe, what burning question haven’t I asked that you are eager to share with our biotech scientists?

Giuseppe Licari [00:11:12]:
Well, maybe “what comes next” in this field. That could be a burning question. My answer would be that with new machines, GPUs, and computational power increasing continuously, in the future we’ll be able to simulate bigger systems for longer periods.

I think simulations will eventually reproduce nearly any step in the development space, supporting more and more phases in drug development. With increasing computational power, we’ll be able to do more and more. I’m really looking forward to seeing how this field evolves in the years to come.

David Brühlmann [00:12:02]:
Giuseppe, what is the most important takeaway from our conversation?

Giuseppe Licari [00:12:08]:
I’d say that in silico methods have been around for many years and are now standard tools to support work across drug discovery and development. They’re here to stay. My message is: don’t be afraid to explore them, be curious, and use them when they’re helpful.

David Brühlmann [00:12:44]:
Yes, scientists, why not use these technologies? Giuseppe, where can people get a hold of you?

Giuseppe Licari [00:12:52]:
LinkedIn is the easiest way. People can search for my name, and I’m happy to exchange with anyone curious about these techniques.

David Brühlmann [00:13:03]:
Excellent. Smart Biotech scientists, please reach out to Giuseppe to exchange on in silico approaches. Once again, Giuseppe, it’s been fantastic. Thank you so much for being on the show today.

Giuseppe Licari [00:13:15]:
Thanks to you, David, for what you do and for the invitation. It was a pleasure for me as well. Thank you.

David Brühlmann [00:13:23]:
What a masterclass in computational formulation development. Giuseppe has given us a roadmap from theory to practice, showing how in silico approaches are becoming indispensable tools in the biotech scientist’s arsenal.

If these insights resonated with you, take 30 seconds to leave a review on Apple Podcasts or wherever you’re listening. Your feedback helps us bring more expert conversations to the biotech community.

Thank you for leaving a review, and thank you for tuning in. Until next time, keep making bioprocessing smarter, one innovation at a time.

Smart scientists, that’s all for today on the Smart Biotech Scientist Podcast. Thank you for joining us on your journey to bioprocess mastery. If you enjoyed this episode, please leave a review on Apple Podcasts or your favorite platform. By doing so, you help empower more scientists like you.

For additional bioprocessing tips, visit www.bruehlmann-consulting.com. Stay tuned for more inspiring biotech insights in our next episode. Until then, let’s continue to smarten up biotech.

Disclaimer: This transcript was generated with the assistance of artificial intelligence. While efforts have been made to ensure accuracy, it may contain errors, omissions, or misinterpretations. The text has been lightly edited and optimized for readability and flow. Please do not rely on it as a verbatim record.

Next Step

Book a free consultation to help you get started on any questions you may have about bioprocess development: https://bruehlmann-consulting.com/call

About Giuseppe Licari

Giuseppe Licari has served as a Principal Scientist in the Computational Structural Biology group at Merck KGaA since 2022, where he helps design and implement digital tools to analyze biotherapeutic molecules. His work includes studying how various excipients contribute to protein stabilization, with the goal of informing and improving formulation development.

Before his time at Merck, Giuseppe worked at Boehringer Ingelheim, where he helped establish computational methodologies for assessing developability and forecasting protein behavior through in silico modeling.

He completed his PhD in Physical Chemistry at the University of Geneva in 2018, followed by a postdoctoral role in the Theoretical and Computational Biophysics Group at the University of Illinois at Urbana–Champaign, focusing on molecular simulations of proteins interacting with biological membranes.

Connect with Giuseppe Licari on LinkedIn.

He is also a biotech technology innovation coach, technology transfer leader, and host of the Smart Biotech Scientist podcast—the go-to podcast for biotech scientists who want to master biopharma CMC development and biomanufacturing.  

Hear It From The Horse’s Mouth

Want to listen to the full interview? Go to Smart Biotech Scientist Podcast. 

Want to hear more? Do visit the podcast page and check out other episodes. 
Do you wish to simplify your biologics drug development project? Contact Us