Having just 2 weeks left of my 3-month secondment at Johnson & Johnson, done within the ALLODD Marie-Curie Network, it’s important to reflect on it and express my gratitude to everyone involved.

My main focus at Johnson & Johnson was combining active‑learning AI with binding free‑energy calculations. Instead of running expensive simulations on every compound, one can let the AI choose the ~10% of molecules, run rigorous GPU-based free‑energy calculations on those, and use ML to fill in the rest. The results seem to be quite similar to the explicit free-energy calculations of the entire dataset, at a fraction of computer time!

Big thanks to Dr. Vytautas Gapsys for directly supervising me during this internship, and to Dr. Vineet Pande and Dr. Gary Tresadern for their guidance. This mix of physics-based methods and AI is a real industrial state-of-the-art.
​
Overall, the experience of working in a real industry in a computational drug discovery team was very enriching, especially for a PhD student from academia. 

Almost four years ago, when I started looking for PhD opportunities, I didn’t even know what a Marie Curie fellow was. I had just graduated with a Master’s degree in Bulgaria, eager to continue in science and work abroad - but I had no idea how to make that happen. I knew I wanted to enter the field of Drug Discovery and gain expertise that would allow me to contribute to bringing novel medicine to people in need.

To do so, I decided to pursue a PhD in this area and I began searching for positions in Bulgaria and abroad. That’s when I discovered the EURAXESS website - a platform that promotes PhD opportunities across Europe. To my surprise, just a few months into my search, several positions appeared as part of a newly funded Marie Curie ITN project called AlloDD - Allostery in Drug Discovery. I didn’t know what an ITN was, or even what allostery meant, but many of the positions were related to Computer-Aided Drug Discovery (CADD) and required a background in computational chemistry - literally perfect for my profile. Naturally, I started reading the project description, learning about the ideas and goals behind it, and soon after, I liked it so much that I ended up applying to nearly half of the available positions.

The idea behind an ITN is in its name: an Innovative Training Network, funded by the European Union - a project developed by high-profile principal investigators (PIs) with deep expertise in a specific topic, who are looking for early-stage researchers (ESRs) to train and mentor into a next generation of experts. So, it turned out an ITN was exactly what I was looking for in a PhD - but I had no idea just how much more it could offer.

A few months after applying, I was lucky enough to be selected as AlloDD’s ESR3, to become a PhD student under the supervision of Dr. Xavier Barril and Dr. Jordi Juarez. I moved to Barcelona, and my ESR/PhD journey began.

The AlloDD project turned out to be a true ITN in every sense. Fourteen ESRs, fourteen PIs, across twelve countries—spread across Europe like a real network. The project included multiple activities designed to help us train in the field of allostery and connect with knowledgeable scientists. Our interactions grew each year through network meetings and planned secondments – secondments that provided valuable training but also opportunities for new collaborations, and meetings, both in-person and online, which equipped us with the knowledge and transferable skills necessary for any career path in science.

As the years passed, the network became even stronger. All of us connected and got involved in each other’s work – it was like having fourteen PhD projects instead of one – such an exceptional experience. Not only the ESRs but also all the PIs closely monitored our research progress and offered invaluable advice and guidance. After three years of this dynamic and insightful environment, by the end of the project, we – ESRs, PIs and partners, the whole network had become like a family. A family of scientists who support each other, teach each other, share knowledge and grow together.

For me, being part of the AlloDD ITN has been the most precious experience. I went from someone who didn’t know what allostery or CADD was, to a scientist with a deep understanding in these fields and proficient in a variety of techniques applied in the discovery of new drugs. I traveled near and far for secondments, network meetings, and conferences. It is worth mentioning that the project provided secure funding, which enabled all of these enriching activities and speeded up my research and scientific development. But most importantly, I had the chance to connect with so many brilliant people: the PIs, who are leaders in their fields, and the ESRs, all bright minds with bright futures. I feel incredibly honored to have met them all at the start of my career, and lucky to call them my network. Now, after 4 years, I can truly say that applying for the AlloDD ITN was the best decision of my life.
​
P.S. If you’re reading this and considering doing a PhD abroad, I would advise you to apply for a Marie-Curie ITN. And I only hope all ITN projects are as well-planned, well-organized and well-executed as the AlloDD.

​Researchers from MIT, Valence Labs, Recursion, and ETH Zurich have developed Boltz-2, a foundation model that advances both biomolecular structure prediction and binding affinity estimation. This work addresses a longstanding computational challenge in drug discovery: accurately predicting how tightly small molecules bind to protein targets.
 
Current methods for binding affinity prediction face significant limitations. Free energy perturbation (FEP) simulations provide high accuracy but require substantial computational resources, often taking days to evaluate small compound sets. Faster approaches like molecular docking can screen large libraries quickly but lack the precision needed for reliable drug development decisions. Boltz-2 attempts to bridge this performance-speed gap.
 
The model builds on previous structure prediction advances like AlphaFold3 and Boltz-1, incorporating several key innovations. The training dataset combines experimental structures from the Protein Data Bank with molecular dynamics ensembles, exposing the model to both static equilibrium states and dynamic fluctuations. The researchers curated millions of binding affinity measurements from public databases, standardizing diverse experimental protocols and filtering for data quality. The architecture includes specialized components for both structure prediction and affinity estimation, with the affinity module operating on the model's structural representations.
 
On the FEP+ benchmark, Boltz-2 achieved correlation coefficients approaching those of FEP methods while running over 1000 times faster. In the CASP16 affinity challenge, it outperformed all submitted entries without specialized tuning. The model also demonstrated practical utility in virtual screening experiments, successfully identifying high-affinity binders for the TYK2 kinase target when validated against absolute binding free energy calculations.
 
The researchers acknowledge several limitations, including variable performance across different protein families and challenges with large conformational changes upon binding. They note that accurate structure prediction remains a prerequisite for reliable affinity estimation.
 
Boltz-2's code, model weights, and training data are being released under an open license, providing the scientific community with access to both the trained model and the complete training pipeline for further development and application.

In April 2022, 9 young scientists met in Vienna. They all had recently moved to a new country and were trying to establish a new life there, clueless about what was waiting for them in this new, exciting journey they embarked on. From different countries and different backgrounds, but they all had one thing in common: Their passion for science, which united them, over and over again, in different countries throughout the following years.

New faces joined them in Barcelona. As they learned more about each other’s projects in detail, they also discovered more about different cultures, languages, and perspectives. Then they welcomed new members in Strasbourg, and the missing piece of the puzzle was found.

In the ALLODD events, during their secondments in each other’s labs, and via social media, they got to know each other better. They were going through the same problems after all: Not getting the anticipated results in their research, dealing with bureaucracy, and feeling lonely sometimes. They supported each other and occasionally had venting sessions to let it all out.

They advanced in their research projects, grew as people and grew older inevitably. Some entered a new phase of life, some said farewell to their 20s, some got married, and some are getting prepared to welcome parenthood. They are not the same people as they were in 2022. Later on, some couldn’t make it to some ALLODD events, but never drifted apart, because they were still connected in so many ways, they cared about each other.

When they met for the last time in Berlin, they promised to keep in touch and not to be strangers.

…

Having been a part of such a project was such a blast
I remember the beginning so clearly but then it went too fast
I met brilliant scientist but more importantly great friends
Hope that we stay in touch even when this whole thing ends
No matter where we end up and if we go where the wind blows
Many thanks for being there for me through highs and lows
Standing right beside me on my happiest day,
Or supporting me wholeheartedly from afar in a virtual way,
I am grateful for all the amazing conversations we held,
For the exciting adventures and fond memories we had,
In the near future, let’s take a plane or catch a train
And meet sometime somewhere in the world again

Last week, I found myself surrounded by nearly 100 early-career researchers at a doctoral conference we organized with the topic "beyond the defense" in mind. The energy was infectious—brilliant minds presenting their research, from novel cancer therapeutics to diagnostic tools. But as the day progressed, I noticed a pattern in the conversations during coffee breaks. The most common question wasn't about methodologies or results—it was "What comes next?"

As PhD students, we spend years mastering the art of discovery, meticulously designing experiments and pushing the boundaries of human knowledge. But rarely do we learn what happens when our discoveries need to become real-world solutions that actually help patients. That gap between the bench and bedside? It's where most academic breakthroughs go to die.

This question hit particularly close to home because I've been asking it myself. As I navigate my own PhD journey in drug discovery through the Marie Curie ALLODD network, I've become increasingly curious about the innovation ecosystem—the bridge between academic labs and patient impact. So when an opportunity arose to explore medical technology valuation through a unique collaboration between Karolinska Institute and the University of Minnesota's Carlson School of Management, I jumped at it. What I discovered completely changed how I think about innovation.

When Scientists Meet MBAs: A Meeting of Minds
The course format was brilliantly simple: put PhD students from KI together with MBA students from Carlson, give them real medical technology assessment challenges, and watch what happens when two completely different worldviews collide.
On one side, you had us—the scientists. We spoke in terms of mechanism of action, clinical endpoints, and regulatory pathways. We could dissect a drug's molecular target with surgical precision but would get lost when asked about market penetration strategies. On the other side were the MBA students, fluent in financial modeling, competitive analysis, and go-to-market strategies, but who might struggle to distinguish between a small molecule and a biologic.
The magic happened in the collision.

Working through technology valuation cases, I watched as my MBA teammates approached our drug discovery research with questions I'd never considered: "What's the total addressable market? Who are the key competitors? What's your intellectual property position? How does reimbursement work in different healthcare systems?" Meanwhile, they were fascinated by our ability to assess technical risk, understand regulatory science, and evaluate whether a proposed mechanism was actually feasible.

One particularly eye-opening moment came when we were evaluating a novel diagnostic technology. I immediately dove into the technical specifications—sensitivity, specificity, and analytical validation requirements. My MBA partner looked at the same technology and asked, "But who's going to pay for this? How does it fit into existing clinical workflows? What's the cost per test?" Both perspectives were essential; neither alone would have led to an accurate assessment.

The course taught us frameworks for technology valuation that combined both lenses: discounted cash flow analysis that accounted for technical risk, real options valuation that considered both scientific and commercial uncertainties, and market assessment that factored in regulatory timelines. But more importantly, it showed us how innovation actually happens—not in isolation, but through collaboration between complementary skill sets.

The Hidden Reality of Innovation
This experience illuminated something crucial about the technology transfer ecosystem that isn't taught in graduate school: successful innovation requires translation, not just discovery.

Most academic discoveries never make it to patients not because the science is bad, but because there's a fundamental communication gap between the worlds of research and business. Scientists are trained to think about statistical significance and mechanistic understanding. Investors think about market size and return on investment. Regulators think about safety and efficacy. Clinicians think about workflow integration and patient outcomes.

These aren't competing priorities—they're all essential pieces of the same puzzle. But too often, they exist in silos.

Universities have technology transfer offices designed to bridge this gap, but the reality is more complex. A typical tech transfer process involves invention disclosure, patent application, market assessment, licensing negotiations, and ongoing relationship management. Each step requires different expertise and different ways of thinking about the same underlying science.

The most successful examples of academic technology transfer happen when teams understand multiple perspectives from the start. Think about the development of CAR-T cell therapy—it required not just immunology expertise, but also manufacturing know-how, regulatory strategy, and business model innovation. The scientists who founded companies like Kite Pharma didn't just make scientific breakthroughs; they learned to speak multiple languages.

The Skills They Don't Teach in Graduate School
Reflecting on the KI-Carlson experience, I realized how many crucial skills are missing from traditional PhD training:

Market Awareness: Understanding not just whether your research could work, but whether anyone would want it and pay for it. This means learning to assess competitive landscapes, understand healthcare economics, and think about adoption barriers.

Financial Literacy: Being able to build basic financial models, understand investment criteria, and communicate value propositions in business terms. You don't need an MBA, but you need to understand how investors think.

Regulatory Intelligence: Knowing how your research fits into approval pathways, what evidence standards apply, and how regulatory requirements shape development strategies. This is especially crucial in drug discovery, where regulatory risk can make or break a program.

Cross-Functional Communication: The ability to translate complex scientific concepts for non-scientific audiences without dumbing them down. This isn't just about making pretty slides—it's about understanding what different stakeholders care about and framing your work accordingly.

Partnership Building: Most innovations succeed through collaboration, not heroic individual efforts. Learning to identify complementary expertise and build productive working relationships across disciplines is essential.

Your Research, Your Future: Practical Next Steps
So what does this mean for you as a PhD student? Start by honestly evaluating your research's innovation potential:

• Assess the landscape: Who else is working on similar problems? What approaches have failed and why? Where are the white spaces?
• Understand your value proposition: What makes your approach unique? What problems does it solve that current solutions don't address?
• Map the pathway: What would it take to get from your current research to a real-world application? What are the key technical, regulatory, and commercial hurdles?
• Build your network: Connect with your university's technology transfer office. Attend industry conferences. Find mentors who've made the transition from academia to industry or entrepreneurship.

Consider seeking out experiences like the KI-Carlson program. Many universities offer innovation and entrepreneurship courses designed for scientists. Organizations like AAAS, NIH, and various industry associations provide workshops on technology transfer and commercialization.

Most importantly, start thinking about innovation early in your PhD, not as an afterthought. The decisions you make about research direction, intellectual property, and collaboration can significantly impact the ultimate real-world potential of your work.

Bridging Two Worlds
That room of 100 eager PhD students represents incredible untapped potential—not just for scientific discovery, but for innovation that changes lives. Each person there is working on research that could potentially help patients, improve healthcare, or solve pressing global challenges. But potential alone isn't enough.

The KI-Carlson collaboration taught me that innovation isn't just about having great ideas; it's about building bridges between different worlds of expertise. It's about learning to see your research through multiple lenses and finding collaborators who complement your strengths.

Understanding the path from bench to bedside isn't just about career options—though it certainly opens doors to industry, consulting, venture capital, and entrepreneurship. It's about maximizing the impact of all those late nights in the lab, all those failed experiments that taught you something new, and all that passion for discovery that got you into science in the first place.
​
Because at the end of the day, the goal isn't just to publish papers or graduate with a PhD. It's to contribute to human knowledge in ways that make the world a little bit better. And sometimes, that requires learning to speak more than one language.