Skip to main content
European Commission logo print header

Automated computational design of site-targeted repertoires of camelid antibodies

Periodic Reporting for period 2 - AutoCAb (Automated computational design of site-targeted repertoires of camelid antibodies)

Reporting period: 2020-07-01 to 2021-12-31

Dozens of antibodies are in routine clinical use for treating life-threatening diseases. The conventional route to antibody discovery and optimization relies on animal immunization and lab selection of improved variants. These methods are powerful, and yet, they ultimately rely on trial-and-error processes. The reason we still rely on such processes despite their low scalability is a lack of control over biomolecular recognition. The AutoCAb project is dedicated to developing a new computational strategy for designing effective antibodies while increasing our understanding and control over biomolecular recognition. The AutoCAb project develops cutting-edge computational and experimental methods that allow designing, for the first time, millions of antibodies that target a specific site on a molecule of interest and synthesizing all of these antibodies accurately and economically. Finally, selection experiments followed by next-generation sequencing allow us to monitor which designs bind their targets and to use advanced machine-learning methods to infer rules that would improve the chances of success in the next round of experiments. At the conclusion of AutoCAb, we expect to have a suite of methods for generating repertoires of functional proteins -- not just antibodies but also enzymes. Such methods will help us address some of the most pressing societal problems, including quickly designing antibodies that target viral or disease-causing proteins and enzymes to degrade new-to-nature pollutants or produce fuels from biomass.
We have established the computational pipeline for designing repertoires comprising millions of specific nanobodies. We have also established an economical and practical approach for synthesizing such a large and diverse repertoire and have applied this approach to several unrelated target antigens. For one of these antigens (a disease-related receptor protein), we have experimental data showing medium-affinity binding. We are now characterizing this binder and optimizing it further. In parallel, we have found that our design approach often results in non-specific binders and we are developing a second generation of our design method to address this problem. We have so far published new methods for designing protein backbones in antibodies and enzymes and are writing publications on other design methods and approaches for large-scale DNA synthesis. These methods will enable the design of large repertoires of new binders and enzymes. We have also developed and published methods for improving the stability of challenging proteins with therapeutic potential, including enzymes that can address antibiotic resistance.
The AutoCAb project is the first of its kind to design massive repertoires comprising millions of functional proteins. To enable this strategy, we have already developed new and automated methods for design and DNA synthesis and have applied them to proteins other than antibodies, demonstrating their generality. An important goal of AutoCAb is to learn protein design principles directly from the experimental data on the millions of designed antibodies. We anticipate that this would lead to a very fast and massive improvement in protein design capabilities. In addition, the methods we develop can also be applied to design superior proteins through considerably less experimental effort than other methods. We expect that these methods will increase the scale and complexity of challenges that can be solved using protein design methodologies, including the design of hyper-stable enzymes for green chemistry, biosensors to detect specific molecules in patient blood samples, diagnostics and therapeutics.
Artistic rendition of constructing a new functional protein from diverse backbones