DDW Editor Reece Armstrong sits down with Tarek Samad, Lundbeck’s Senior Vice President and Head of Research to discuss AI and supercomputing within the drug discovery industry following the company’s recent agreement with the Danish Centre for AI Innovation (DCAI).

In May this year, Lundbeck announced an agreement with the DCAI in which it would be able to run Gefion, Denmark’s flagship AI supercomputer with the goal to accelerate drug discovery and development in the fields of neurology and psychiatric disorders. 

With such focus being placed on advanced technologies in pharma in recent years, we began the conversation discussing the current position of AI and supercomputing within the industry and where the potential lies. 

“I can tell you, confidently that AI is used at every step, in the process of drug discovery. From ideation to screening, to running clinical trials to reaching patients to data analysis,” Samad says.

“It’s used at every step. There are pockets where the use is higher, with more examples of success, like in early drug discovery and screening. But it’s used throughout the value chain in the pharma industry and the biotech industry as well.” For Lundbeck, the recent partnership with DCAI is a step by the company to “double down on the use of AI and full computational power in drug discovery”, Samad explains. 

What this means specifically is that Lundbeck will use the computing power of the Gefion supercomputer to derive greater insights from its own AI algorithm in relation to molecule discovery and drug optimisation within drug discovery and development. 

In terms of the project’s impact, Samad hopes that there will be two main outcomes. The first is based on the scope of the company to target solid biology in previously undruggable areas. The second impact is pace.

“This partnership, provides, I would say relatively easy access for our scientists to a high GPU computational power and capacity to fully use our internal AI platform research,” Samad says.

For Lundbeck’s first project using Gefion, Samad explains that the company aims to speed up drug candidate selection by approximately five times and explore up to 1,000 times more molecules compared with what standard AI computers can currently do. 

Currently, the abundance of data that is generated by AI research projects in pharma present a number of obstacles for the industry to overcome. 

“With the sheer amount of data you generate with AI, the bottleneck becomes computing power and GPU capacity,” Samad explains. 

“For example, what we refer to as in silico screening of ultra large libraries of potential drug candidates, this was done manually in wet labs where you would screen 300, 400,000 compounds and it would take several months.  

“Now you multiply this by several orders of magnitude and you get an idea of the number of different structures of molecules that you can test at the same time. So you need that high computing power.

Samad continues by saying that having access to Gefion will allow Lundbeck to utilise its own AI-driven algorithm to both find the right drug candidates and also understand the molecular structure of targets. 

“If we understand the structure of our target, we are more able to design the right molecule with the right structural but also functional properties to increase the chances of identifying drug candidates for a given target for an unmet needed disease in a shorter time.”

What this means for Lundbeck is that the company can theoretically identify candidates using AI for its predictive validation of molecules. This lets the company understand which molecules are predicted to be efficacious before it can verify those molecules with select wet-lab experiments. 

Through the partnership, Lundbeck hopes it can better tackle neurological health, an area that has historically been difficult to treat and which as a result, has seen the pharma industry step back from tackling the unmet need in this area. 

This unmet need whether it be mental health, neurodegenerative or neurological disorders, as Samad says, isn’t going anywhere.

“There’s a very high societal burden, both material burden, but also health burden. Not only on patients themselves,  but on caregivers and families. It affects society ultimately. It affects the GDP of the country,” Samad says.

It’s something Samad says that the industry can’t turn its back on. 

The good news is that neuroscience research is on the rise. In 2024, manuscript submissions to the Neuroscience journal increased by 64.6% compared to 2023¹.

Samad believes that “technology has been pivotal in the, if you will, the revival of neuroscience.”

“The revival in interest has been primarily propelled by advances in biology, and also advances in an area we call modality.

More often than not, we understand what the root cause of a neurological disorder is, but we cannot design a small molecule to go and tackle the target. Now with the explosion of the different modalities such as RNA therapeutics, biologics, small molecules, we can tackle those.” 

These new and emerging modalities which can be applied to neuroscience research, draw stark comparisons to the advances we’ve seen happen in oncology treatments – which are now more advanced and targeted than ever before.

Samad agrees with the comparison, saying how neuroscience used to be area that was restricted only to small molecules due to their ability to penetrate the blood brain barrier (BBB) without any modification or transport. 

Now, biologics are making an impact in neuroscience research and bringing more treatments to patients. Researchers are finding workarounds to the fact that large molecules like biologics cannot cross the BBB, by pairing them with delivery systems such as antibody-drug conjugates (ADCs). 

Market approvals in diseases like Alzheimer’s for biologics have helped drive success in neuroscience and this is being followed by other modalities such as RNA therapeutics according to Samad.  

“RNA therapeutics is following. So that is a clear driver. Not the only one, but it is a clear driver. Knowledge of biology and understanding disease pathogenesis is another driver as well. And now we understand it better due to genetics, due to availability of human data, human biology data, and the ability of us to collect all these data with AI and having the predictive possibility to look at it too,” Samad says. 

Continuing on the topic of neuroscience, the conversation shifts to the unmet need in mental health treatments and the need for more novel approaches. 

For Lundbeck, the company’s focus on neuroscience means it has a deeply defined understanding of how novel mechanisms in biology can be utilised to bring better treatments to patients. 

The challenge in neuropsychiatry is bringing differentiated molecules through to market. 

“So, we look at the biology in a very rigorous way and only when the mechanism is novel and well validated is where we pursue neuropsychiatry [assets]. And you don’t find a lot of novel mechanisms in neuropsychiatry.

So that is the reason why you may find fewer assets in neuropsychiatry. Not because we are not in neuropsychiatry but because the bar is high.”  

We end the conversation discussing the advantages and strengths of Denmark’s ecosystem for AI and biotech.  

It’s a topic that Samad touches on regularly with stakeholders and partners, referencing how successful biology has often emerged from pharmaceutical and biotechnology hubs around the world, such as Boston. 

“What are the magic ingredients, if you will? You look at excellent academic institutions, you look at capital, right? There’s capital and in presence of pharma and biotech and entrepreneurs.”

These ingredients are found at various stages in the Copenhagen area, with capital being propelled by initiatives like the Novo Nordisk Foundation and the Lundbeck Foundation, alongside the city’s pharmaceutical presence. 

“You’ve got excellent academic institutions, absolutely amazing universities in Copenhagen. You’ve got a lot of entrepreneurs, lots of scientists and lots of startups. So, what you’re seeing is an ecosystem that is very conducive for innovation.”

From DDW Volume 26 – Issue 3, Summer 2025 Read the digital issue

Reference:

1. https://ibro.org/neuroscience-is-on-the-rise/#:~:text=2024%20has%20been%20a%20landmark,research%20from%20across%20the%20globe.

Biography:

Tarek Samad is Senior Vice President and Global Head of Research at Lundbeck with over two decades of experience in academia and industry, leading small molecule and antibody biologic programs into the clinic. Prior to joining Lundbeck, Samad was the chief scientific officer at Immunitas Therapeutics, and before that, Head of Multiple Sclerosis and Neuroimmunology Research at Sanofi.