Cyclica successfully integrates DeepMind’s AlphaFold 2

AlphaFold 2 has provided structures for a large percentage of the proteome, perfectly complementary to Cyclica’s AI-augmented platform thus allowing for Cyclica’s drug prediction domains to be able to identify drug binding to all proteins in the human body.

Two weeks ago, Google’s DeepMind made available the source code of AlphaFold 2, as well as a compilation of protein structures of 20 species generated in collaboration with EMBL. At the time, Andreas Windemuth, Cyclica’s  CSO, and I wrote a blog post where we highlighted the impact this could have on scientific/research advancement and referenced a few things that excited us at Cyclica. 

In this post, I’d like to share some exciting and potentially groundbreaking work we’ve undertaken. Over the past couple of weeks, my incredible team at Cyclica has worked diligently to add AlphaFold 2 library into our MatchMaker screening pipeline. As a quick refresher, MatchMaker is our deep-learning engine that uses structural and experimental assay data to predict drug-target interaction proteome-wide, and powers both Ligand Design for multi-objective and multi-targeted drug design, and Ligand Express for off-target profiling and phenotypic screening target deconvolution

First and foremost, the library is fantastic, and the structure files behave very well - thank you, DeepMind and EMBL! We now have ~200,000 non-redundant pockets from 17,712 proteins, roughly doubling the coverage in our database. This represents the largest AI-augmented proteome-wide screening capability in the pharma industry. In just under a week, we identified potential binders for targets of interest by evaluating across 17,712 proteins. With Ligand Express, we actively prioritize the programs for those with the preferred off-target profile and ADMET properties. Once our team of drug hunters evaluate this data, we will ideate and advance hundreds of drug discovery programs over the next few years across a range of therapeutic areas, namely those with high unmet medical needs. This will lead to the largest and most diversified portfolio of assets in the pharma industry - all part of our vision of advancing molecules to medicines by accelerating the biotech pipeline of the future. 

How do we do this? Well, first, unlike other methods, MatchMaker takes a coarse-grained view of protein structure, which makes it ideal for taking advantage of modeled protein structures. Ligand-based QSAR models, for example, don’t use protein structure at all, while target-centric virtual screening methods require detailed information at an atomic level of detail that even AlphaFold 2 is hard-pressed to provide. Second, unlike other methods, with MatchMaker, we are able to conduct a drug design campaign against the entire proteome in record time. What would take other techniques - both empirical and computational - years to accomplish took us only a matter of hours. Last, the availability of whole structural proteomes for multiple species provides MatchMaker with the ability to work with model organisms and pathogens to widen its scope in human health. Beyond that, infectious disease, animal health and agriculture now become opportunities for us to rapidly have a greater impact. DeepMind and EMBL are providing 20 species now, thousands are planned for later, and the open-source enables us to address any species on demand. 

We will share more about the impact of AlphaFold2 on drug discovery using Cyclica’s workflow during the next few weeks. Stay tuned !!

Naheed Kurji, President and CEO
Dr. Andreas Windemuth, Chief Science Officer

Naheed Kurji, Chief Executive Officer

Naheed Kurji, Chief Executive Officer

Naheed Kurji is the Co-Founder, President and CEO of Cyclica. Naheed is passionate about building AI-augmented technologies that enable researchers to make more strategic and informed decisions in Healthcare and the life sciences. He spends the majority of his time obsessing over Cyclica’s culture, defining its strategy to best effect change in the pharma industry to achieve the company’s vision, and exploring opportunities for continued innovation.

Related Posts

Using Cyclica’s technology to identify repurposed drug candidates for COVID-19

At the time of writing this article, over 1.3 million people have been confirmed to be infected...


A look at yesterday, today, and tomorrow

Cyclica has come out of its best year yet, with a clear vision for the future and focused...


Building a neo-biotech: It's all in the (dream) team. Meet them

I’m often asked how I can lead a deep tech biotech company that’s at the intersection of AI,...