AION Labs, a first-of-its-kind innovation lab focused on AI technologies and computational science to solve therapeutic challenges, recently announced its official launch and the opening of its international headquarters.

AION Labs was formed through the alliance of four leading pharmaceutical companies – AstraZeneca, Merck, Pfizer and Teva – and two leaders in the hi-tech and biotech investment sphere , respectively – Amazon Web Services Inc. (AWS) and Israel Biotech Fund (IBF). These giants in tech and pharma are working in concert to develop groundbreaking AI and computational innovations aimed to leverage the cloud to transform how new therapies are discovered and developed.

DocWire News sat down with Mati Gill, CEO of AION Labs, and Dr. Yair Benita, CTO of AION Labs, to discuss how the company plans to develop new technologies that meet the challenges of working in the pharmaceutical industry. See what they had to say.

DocWire News: Can you tell us about yourselves, and about AION Labs?

Mati Gill: Thank you. So, as you mentioned, my name is Mati and I’m an Israeli American. I was born in the US, but I’ve been living in Israel. This startup nation, which is widely known as such for several decades, a veteran of the industry was working at Israel’s largest life sciences company called Teva Pharmaceuticals for a little over a decade, and very excited to now lead up this new innovation lab and to help bring some hope to both the industry and to humanity entirety, leading up this alliance of our six partners plus BioMed X, namely Amazon, Israel Biotech Fund, Pfizer, AstraZeneca, Merck, and Teva Pharmaceuticals together with BioMed X, our R&D engine that we brought in from Germany. It’s very exciting for me to be able to lead up this new consortium and help to build and invest in the next generation of AI and computational based startups that will come out of our lab and hopefully grow in Israel successfully, and thankful that our Chief Technology Officer, Yair Benita, decided to join. He’s my partner in this ride, and he can introduce himself.

Yair Benita: Thanks, Mati. Hi, pleasure to be here. I’m Yair Benita, I have a background in pharmacology and computer science. I drifted from the biological field into the computer science field around 2000 when the genome was sequenced. Was very passionate learning more about it, trying to understand the data and how it can impact drug discovery. I’ve been in this interface of biology and computer science ever since, and my true passion is to use data to drive and improve drug discovery and development. That’s what I’ve been doing throughout my PhD, then post-doc in Boston, worked for Merck for a few years in Boston, and came back to Israel to lead the computational groups in biotech environments. I recently joined. I feel that this is a unique opportunity because of our model and the partnership, which I think we’re going to talk about more in this interview.

Talk to us about the partnerships that were formed to launch AION Labs.

Mati Gill: Sure. Maybe I could just start a little a bit by how this came to be about. Israel, as I mentioned before, is really a hotbed for innovation in several different fields, and has a longstanding reputation and track record for early-stage research in the life sciences sector, where nine blockbuster drugs have come out of Israeli academia and research. However, the advanced development and the advanced stages of the development of manufacturing, commercialization of drugs, with the exception of Teva Pharmaceuticals, has not been as prevalent as would be expected from a country with thousands of post-docs in the life sciences sector and leading academic institutions that have churned that IP, mostly because the advanced stages after the early-stage research happen outside of Israel.

So the Israeli government wanted to put together a strategy and a program where they could really accelerate the industry here and help jumpstart the industry and build an ecosystem for advanced development and for companies to really continue the development of ideas in the pharmaceutical space. However, they understood that if they were to focus on the current activities of the pharma sector, Israel would not be competitive because we kind of missed the boat on that, from a government standpoint and a national standpoint. So in order to do so, the Israeli government put together a strategy that they call bio convergence, meaning that they are mixing and trying to be bring together the capabilities on the engineering side, the mathematical computational side, including artificial intelligence, machine learning, comp bio, et cetera, together with classic biology and how they can really help revolutionize the way that we discover and develop new drugs in the pharma space.

That strategy brought a competitive tender together to establish a calling on global companies to come and establish a new innovation lab with the support of the Israeli Innovation Authority, the Israeli government arm that helps fund innovation, to come together and build an innovation lab with an infrastructure inside that has both a wet lab and a computational lab, where startups can either be formed or invested in from a very early stage to do things the right way and taking these ideas from the academia and from the early stage of the lab and become developed, validated, and industrialized. And that’s what we intend to do.

We, fortunately, have multiple partners that decided to join this together. Even competitors collaborating in this space because they feel passionately about the potential that AI has in drug discovery and development, but also that working together would increase the chances for them to succeed, and learning from each other would increase the chances to succeed in a field that the potential has remained untapped for the most part and there’s more to be learned than has actually been really up until now.

In addition to that, they also wanted to work in a very multidisciplinary manner, so we have four elements of our partnership. We have the global pharma companies that came together to be partners in establishing a lab. We have a high-tech partner with a vast experience of computational expertise in the life sciences sector in Amazon web services. We have a venture capital partner, which brings in the business and investment expertise into the stage to help our startups grow from the very early stages until they become growth companies, as well as, of course, our R&D engine of BioMed X, which has been working in a pharma space with R&Ds of the global pharma to decipher what the challenges are that we’re going to focus on in that sense. And that’s how we’re going to work, where we’re going to take, with the help of BioMed X, the top R&D challenges and research challenges that remain untapped from the global pharma and take them for a global call for applications and find the best talent that want to come and become scientists, founders in our lab to help establish new ventures.

What are the challenges in the health care space that AION Labs is looking to solve?

Yair Benita: So, first we have a process where we go to through each of the pharma partner companies and identify the challenges that can make a real impact, where a computational technology is suitable and AI would be suitable. Then, in this process, we actually aggregate all the challenges from all the pharmas, and then either merge some of that are similar, but come up with the top few that all pharma partners think are good, or at least a few of them. And that defines our problem space.

The first one that was identified and defined was computational prediction of antibody structure or antibody sequence. And the way it works, if you think about it currently, what pharma companies do to develop antibodies are either screening through, for example, the so-called phage libraries, or immunization in animals, where you immunize with the antigen and then isolate, try to identify the antibody is the animals developed. Both approaches are very laborious or take a long time. It takes 2 to 3 years to develop an antibody, and we’re looking for a way to improve on that process.

Ideally now, the challenge is if we provide the three-dimensional structure of a protein, can we predict the antibody sequence computationally and then go generate it. Even if we come up with a hundred different candidates and one of them would be a great antibody, it would be a huge success. We feel that the time is right, because there’s been quite a lot of advances in that field of protein folding, and three-dimensional structures. It’s obviously a very big challenge, but it’s also very big in the sense that it’s actually difficult to, to perform, but actually it will have a very big impact. It’s also a challenge that is considered pre-competitive. It’s a technology that could be available to all pharmas, not only our partners, and will advance the field significantly if we’re successful. So these are the types of challenges that we’re expecting that will push very innovative, very ambitious, with great impact.

How can AI technology and computational science be used to enhance drug discovery and development?

Yair Benita: Again, we have to find the problems where, first, we have the data which can be used, both for learning and for testing, and that data is large enough. The problem has to be a problem that fits. For example, a three-dimensional structure of a protein is a problem that fits a computational technology, or, for example, taking a given antibody and improving it so it can be a better drug in terms of its attributes like solubility, viscosity, pharmacokinetics, pharmacodynamics. All of those things. We have a mathematical basis behind it, so it can be used. There’s enough understanding to start building around it. So we’re looking for such problems, and these problems are scattered around the drug discovery and development process. But every time, if you think about it, every time we can replace the components that is very currently low-tech and tedious and very expensive with a better AI-driven solution, and that way we envision that will improve the entire process over time.

Do you take a hands-on approach in running AION Labs?

Mati Gill: Well, right now we’re in establishment phase, and like any startup, in establishment phase, I believe you have to take on a hands-on approach, of course, where things are focused on what the matters are that are critical. And then with the help of the team that we’ve built, letting each one of them really lead in a hands-on approach in their field and their structure. But, ultimately, you have to be hands-on in the areas that are most critical for the business.

What’s the ultimate goal of AION Labs, and where do you see the company in three years?

Mati Gill: In 3 years from now, hopefully we will have either created or invested in at least 15 new startups that will have gone through our program. And 3 years from now, already have established at least some startups that have already started to validate their technologies, showing that computational bio can really work in the pharma drug discovery and development phases. With the help of our partners, we’re hoping to be able to help show that we can bring innovation and disruption to the way that we develop and discover new drugs.

Any closing thoughts?

Yair Benita: I just want to add that one of the goals here is to actually take breakthrough technologies and implement them in pharma, which might sound relatively easy, but it’s actually not very easy. Many companies that developed brilliant algorithms find it difficult to actually convince pharma to use them because it’s hard to validate biologically. And what we’re trying to do is make sure that the technologies are developed together with the pharma partners, and once they’re part of this process, they will adopt it and implement it and test it internally. Because ultimately, validation happens in the clinic on patients, and that has to still be done in conventional ways and cannot be done technologically

Mati Gill: Well, and I think we’re thankful that we have pharma giants, really pharma leaders, that are open to working in new ways.

Yair Benita: Yeah, of course.

Mati Gill: And hopefully we’ll be able to do that with them.

Yair Benita: Yeah. Not only open, but they want to be part of the process. Not only see the end product, because sometimes the end product will not be what they need or want. So they’re going to be part of the process, and I think this is key.