This week Israel-based AION Labs, an AI-enabled drug discovery partnership between global pharma and tech companies like AstraZeneca, Merck, Pfizer, Teva, Israel Biotech Fund and Amazon Web Services, announced the formation of a new startup company dubbed OMEC.AI.
OMEC.AI aims to build a computational platform using AI that can help researchers assess the clinical trial readiness of a drug candidate, identify hidden safety liabilities, and suggest experiments to close any identified gaps.
Gill sat down with MobiHealthNews to discuss OMEC.AI, how the startup came about, and the data it will use within its AI computations to meet its intended results.
MobiHealthNews: Can you tell me about OMEC.AI and its goals?
Gill: Our whole venture creation model is built on three pillars. The first pillar is always starting with a big challenge that, if it was solved, would be truly impactful for patients and, of course, a very strong viable company addressing an industry-needed opportunity.
Secondly, we seek the best scientists and founders to be able to address that challenge with a very lengthy and strenuous in-depth evaluation process.
Third, we set them up as a new startup company with funding, then systematically mentor them for success and support them by giving them the data and everything else necessary for them to be successful.
All of that is conducted not just by the AION Labs team, but really hands-on by our partners in a codevelopment model where everyone works together from day one to help build this company and make it successful for four years.
In this case, OMEC.AI has three supporters from amongst our partners: Pfizer, AstraZeneca and Merck (the German Merck, EMD Serono). Those three are direct investors in OMEC.AI and will have equity in the company, but no IP rights. They each appoint champions from within their R&D organizations to help them and work with them systematically to develop the technology, and have taken part in defining the challenge as well as selecting the candidates.
OMEC.AI is addressing how we take the process of deciding which drug candidate should go into clinical trials, which is probably the most pivotal decision in the pharmaceutical R&D process. Once you decide which drug candidate to bet on as a pharmaceutical or biotech executive, then it goes into a process of investment of hundreds of millions of dollars that you never stop unless the science just fails.
So, really what our partners wanted to do is use artificial intelligence to be able to create a technological platform that would help them make better decisions and ultimately lower the attrition rate and make these drugs safer and more efficient for patients.
The challenge was, how do we take all this data – preclinical data that’s generated plus other sources of data – and create an AI-based platform that would be able to test the drug and tell you what its chances of success are during the clinical trial phases before it goes into humans? And right now, that process is done basically manually with very little technological insights.
Ultimately, we know that the vast majority do not reach approval in the market because they failed at some point during the process. So there’s obviously an unmet need and something that digital and computational technologies should be able to solve if you bring the right people to do that. And that’s what we sought out to find.
MHN: Who are the people you found to set up this team, and what will they be looking to solve?
Gill: They are two artificial intelligence veterans that have worked at the forefront of technology in the AI field in the automotive industry, primarily. They worked at Mobileye, an autonomous driving company that is based out of Israel, but was sold to Intel for $15 billion.
So, they came to us with a technological approach of being able to create a platform that can integrate the data, and in a very ambitious manner that there’s going to be high risk, but also high reward. And our partners love their approach – the R&D partners met with them.
So those three companies, AstraZeneca, Pfizer and Merck, together with Amazon Web Services will support them as well. So they’ll have those four companies, as investors, or supporters and in the case of Amazon, working hand-in-hand with them to be able to develop the technology. We also received a grant from the [Israeli] government to support them. So they receive financing of $2 million, basically, as a pre-seed round. And they’re starting to work this month.
MHN: You mentioned it’s high-risk but high-reward. Can you tell me what some of those risks might be?
Gill: Well, it’s unproven that you can [do this]. To this date, there currently isn’t a technology that can do what they’re trying to do. So, that’s the risk. Can we really create an AI platform with the data accessible to them to be able to ultimately test every type of drug before it goes into clinical trials? To be able to tell pharma, investors or whoever would be a user whether or not this drug candidate has a high probability of success or a high probability of toxicity? That’s something they’re going to try to do. It’s yet to be done. So, therefore, it’s by definition, high risk.
MHN: AI is really only as good as the information that’s put into it. What’s interesting is you have Merck, AstraZeneca, Pfizer and others involved in the project. Is the data coming from each one of these companies? Is there selective data that’s being utilized?
Gill: So, the companies have all committed to providing the data necessary for them to do whatever they need based upon what they have. It’s not like Pfizer is going to say, okay, take all of our historical data without careful selection. The partners are happy to share their data with the startup.
They’re committed to doing that, and they want to do that to help them develop their technology. But they don’t want to share with each other, and they’re not allowed to share with each other, because that would be anticompetitive by definition. So, we’re creating a platform to enable them to be able to share that data in a federated manner in accordance with all best practices.
MHN: What do you hope this program leads to eventually? What do you hope this company can solve?
Gill: What we’re trying to achieve at AION Labs in general, and then specifically for this company, is creating great independent-growth AI-based startups for the biotech field so we can help scientists and researchers – not to replace them, but to really empower them by bringing new technological capability, so that they can optimize the whole process of drug discovery and development.