Antibodies are currently discovered by immunising animals or screening vast libraries of molecules—a slow, expensive process that offers little control over which part of a target they bind to. This project will replace those methods with computer-based design, using artificial intelligence and fragment-based modelling to generate antibodies entirely in silico before testing them in the lab. The problem is that existing antibody discovery technologies cannot reliably target specific regions of a protein, known as epitopes, which is often essential for a drug to work. They also struggle to control properties like stability and solubility, leading to candidates that fail later in development. Computational design bypasses these bottlenecks by letting researchers specify the desired binding site and molecular behaviour from the start. If successful, this approach could slash the time and cost of developing antibody drugs, eliminate the use of animals in discovery, and make it feasible to generate therapeutics against diseases that currently lack good antibody tools. The same technology could also produce better reagents for biomedical research and diagnostics. The project is applied rather than fundamental—it aims to build a working pipeline, not to explore a new principle—but its success would give industry a reliable, cheap route to new drugs.
View original technical description
Because of their ability to bind most molecular targets tightly and specifically, antibodies are increasingly used in biomedical research, diagnostics, and medicine, where they are the fastest-growing class of therapeutics. The rapid surge in the size and quality of structural and biological databases is allowing to introduce innovative computational methods of rational antibody design. The goal of the proposed research is to develop and establish novel computational technologies of antibody discovery and optimisation, by using a multidisciplinary approach that encompasses fragment-based rational design, the development and deployment of artificial intelligence methods, in vitro experimental validation, and in vitro affinity maturation. Rational design at a computer substantially lowers the time and costs required to discover novel antibodies for a target of interest, does not exploit animals, and enables a much better control over the properties of the obtained antibodies. For example, it allows to obtain antibodies binding to specific regions of interest (epitopes) within the target (antigen), which remains a critical challenge with established technologies of antibody discovery, but is of key importance for many applications. Computational design also offers a better control over other properties essential for successful antibody development, including stability and solubility. The proposed research represents a significant step forward towards the establishment of computational design as a competitive technology for the generation of novel antibodies. Computational approaches promise to enable the reliable and inexpensive generation of drugs to combat - and tools to study - many crucial diseases. Overall, the unique opportunities offered by these approaches will enable to address new questions, accelerate discoveries by facilitating experiments, streamline therapeutic antibody development, and provide novel avenues for industry investment.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know