• Sustainability
  • DE&I
  • Pandemic
  • Finance
  • Legal
  • Technology
  • Regulatory
  • Global
  • Pricing
  • Strategy
  • R&D/Clinical Trials
  • Opinion
  • Executive Roundtable
  • Sales & Marketing
  • Executive Profiles
  • Leadership
  • Market Access
  • Patient Engagement
  • Supply Chain
  • Industry Trends

The Case for Pharma, AI, and Israel: Q&A with Mati Gill


The CEO of AION Labs discusses both the new ways that AI is being implemented in drug development and why Israel is becoming the center of AI development in the pharma industry.

Mati Gill

Mati Gill

AION Labs is an Israel-based company with a global reach. CEO Mati Gill spoke with Pharmaceutical Executive about how the company is both utilizing AI in new and exciting ways while also demonstrating to the pharma industry that Israel is the home of AI technology development.

Pharmaceutical Executive: How have you how have you seen AI utilized in this space?
Mati Gill: We are seeing AI becoming more and more relevant. I think that's what made Israel relevant for pharmaceutical investment. AI is more of well-known buzzword, and a lot of times it's actually machine learning systems.

We segmented in four different elements. First is how to use a data centric approach. The fact that we have four pharma companies partnered with us offers us the capability to really leverage additional data, not just from one company. But how do we use a data centric approach to understand the mechanisms in a robust manner that we will then be able to understand? That's really at the foundation, understanding the different mechanisms that are out there.

We know we have one company in that space that supports the pharma companies’ capability to understand disease mechanisms in a much higher manner. Then we have drug discovery, and we have four different companies in that space already. There is one that's doing protein and antibody design. So basically, the function of AI in this case is the capability to actually predict where which body or protein will be able to fold and target various diseases. And, by doing so, be able to address other diseases as well. This would include being able to develop antibodies in a much more precise, robust, and fast manner. The same would be true in the small molecule space, and we have a company in that space as well.

There’s also the molecular space, which is in the protein degradation space (which is a new area of drug discovery). It's still in drug discovery, but it's a little different in the fact that there if you actually bring an antibody, that doesn't mean you actually have a good drug that's going to be able to pass toxicity and efficacy and then be successful in clinical stages. The question is how can we build a one stop shop machine-learning-based mechanism that if you bring the data of an existing antibody, we can make it into a better drug, validate it, and the figure out what lab capabilities are needed to prove that it’s actually a good drug that will be able to ask preclinical and clinical studies.

The third segment is the preclinical stage. There is where you're seeing a lot of regulatory push. We use AI to be able to make animal models more precise, to be able to ultimately replace animal models. That is something that would make sense, because we all know that we've cured cancer in mice many different times, but not yet in humans. The results have generated a lot of excitement from different pharmaceutical endeavors but haven't actually cured cancer yet. We understand that the way a drug is going to behave in preclinical studies is not going to be the same clinical studies ultimately.

Lastly, the fourth segment is the clinical stages. This is where if you actually take a drug and put it in humans as part of bringing it to the approval of the regulatory authorities. Because 90% of drug candidates ultimately do not reach the end, we;re able to bring the capability to learn from a dataset centric approach. Using AI, machine learning capabilities, and advanced models, we believe we can lower that 90% attrition rate to something much less, even if it's 70% or 60%. That's capturing a lot of value.

Being able to bring new drugs home to patients is ultimately what we're doing throughout that whole cycle of understanding diseases, discovering new drugs, and then helping them reach the patient with FDA approval. It’s about being able to try to treat diseases that are currently untreatable or to be able to do it in a much more robust manner.

PE: What makes Israel attractive to the pharmaceutical industry?
Gill: Out of the seven partners, we have VCs that have a base in Israel. And then of course, Pfizer, AstraZeneca, Merck are not in Israeli. I think the reasons why Israel is attractive to the pharma industry is if you look at where the future the industry is headed, there’s a stronger focus on the technology side. It hasn’t always been that way, but that's starting to change.

Pharma companies are understanding that they need to take more of a datacenter approach and actually make it much more equal on the bio and tech side. So, the tech will play an increasingly important role in that aspect, due to tech being the mathematical disciplines of AI and computational technologies. That's where Israel is strong and competitive at a global scale.

Israel has been able to develop ecosystem capabilities that are able to build startups that go on, grow, and contribute to global industries in pretty much every industry with the exception of biotech in the past. Even with the automotive industry, where Israel is not a car manufacturer in any shape or form, we have 600-to-700 startups in the auto tech space. Anyone from Mercedes, Porsche, all the German companies, Japanese companies, and American companies all have presence here in Israel, because they want to be besides the most compelling computational technologies that can differentiate their cars and to go into an integrated system in their cars.

And not one of those startups actually builds a car. What they do is they use AI to make the cars faster, more sustainable, or building autonomous driving. That's what we're seeing now in the biotech space because AI is becoming more relevant. Our industry is one of the last for AI to become very relevant for.

Related Videos