So far in our blog series on Artificial Intelligence in Drug Discovery (AIDD), we looked at active AIDD clusters across the globe, the key techniques used for drug discovery (Part 1 & Part 2), and how various companies are shaping the ecosystem. In this blog, we take a closer look at the role of pharmaceutical companies and their partnerships with AI-driven companies.
Pharmaceutical corporations are the backbone of drug discovery and development. Yet, given the decreasing R&D productivity in the biopharma space, the need to adopt more innovative approaches to drug discovery is clear. Pharma leaders have been proactive and deployed various strategies for adapting to this new environment. They see partnerships as a critical means for success – both with biotech companies with certain disease expertise, as well as those with a sought-after platform. Drug discovery and development is increasingly decentralized, where more and more new ideas and new IP is generated by biotech startups and academic groups across the globe.
Looking at AIDD in particular, the number of partnerships between pharma and AIDD firms has grown from 28 in 2015 to 132 in 2020. All major pharmaceutical companies have announced such partnerships; with Bayer and Pfizer leading on the pharma side and Exscientia, Atomwise, and Cyclica on the AIDD front. Most of these partnerships are structured similar to biotech licensing deals, where IP is licensed against upfront payments, milestones, and royalties.
A case in point is Bayer’s collaboration with Recursion on fibrotic diseases. The deal entails a $30 million upfront payment, plus $100 million for milestones in up to ten drug discovery programs, making the deal potentially worth more than $1 billion. Another example is the partnership between Bristol Myers Squibb and Exscientia, valued at $50 million upfront and up to $125 million in near to mid-term milestones. XtalPi is working with Pfizer to develop a hybrid physics/AI-powered software platform for accurate molecular modeling of drug-like small molecules.
Pharma is also partnering with those providing advanced physics- and modeling-based solutions. Schrödinger is leading this field, establishing collaborations with Ono Pharmaceutical, Sanofi, and Bristol Myers Squibb to find drugs for multiple indications, and forming new ventures such as Bright Angel Therapeutics with MaRS Innovation and Faxian Therapeutics with WuXi AppTec. Nvidia is also entering the space with their physics-enabled AI framework, called Modulus, for drug discovery, which was launched this November.
Quantum computing, an emerging technology, is also gaining increasing attention. Several pharma players have already started working with quantum companies to explore the technology’s capabilities. For example, Boehringer Ingelheim partnered with Google Quantum AI to use quantum computing for molecular dynamics simulations earlier this year. Roche inked a similar agreement with Cambridge Quantum in January this year. In July, Riverlane, Astex Pharmaceuticals, and Rigetti Computing joined forces in a research partnership to drive drug discovery forward with quantum computing.
In parallel to partnering with AI-first players, most if not all pharma companies are also actively and rapidly developing in-house AI capabilities. GSK set up an AI unit in 2017 and established an AI hub in King’s Cross in 2020. AstraZeneca partnered directly with hardware provider NVIDIA and the University of Florida to boost their drug discovery pipeline. Novartis and Microsoft created an AI Innovation Lab to benefit from Microsoft’s AI technologies and Novartis’ life sciences expertise. Sometimes these partnerships even include government or government-related entities, such as when AstraZeneca, Merck, Pfizer, and Teva Pharmaceutical started building AION Labs, an innovation lab focussed on leveraging AI and computational approaches for drug discovery set up in Israel.
Artificial intelligence is revolutionizing drug discovery and development. Biotech and pharma companies will need to reinvent themselves and incorporate AI knowledge into their pipelines – via in-house or partnered approaches. There is great potential for collaboration between these large organizations and the rising stars in the field of AIDD. As the first AI-generated drug candidates start reaching clinical trials, we remain attentive to see how the field evolves.