AI POWERED ANTIBODY ENGINEERING / EVOLUTION
BioDuro-Sundia has established an antibody discovery platform that combines artificial intelligence (AI) and synthetic antibody discovery. The combined platform will enable our biopharma partners to shorten timelines and improve quality of antibody-based drug development programs, enabling faster path to the clinic.
The combined platform can screen tens of billions of high-impact sequences in a high-throughput manner. The sequences are all synthetic, meaning they are purposely designed antibodies with optimized affinity, specificity, stability, and therapeutic developability. The platform is especially useful for difficult targets, such as GPCRs, ion channels, and peptide-MHC complexes, that have consistently failed development as antibody-based therapeutics.
The AI powered antibody technology uses the latest artificial neural network algorithm, machine learning, and molecular dynamics to simulate the directed evolution of each antibody—ultimately generating optimized sequences that specifically and potently bind the target antigen. These AI-directed sequences are quickly advanced into experimental screens. Out of astronomical number of potential sequences, our platform selects the very best subset to advance into screens. A huge amount of AI computation is expended in order to select smaller, but optimal libraries with higher success rate of antibody discovery.
BioDuro-Sundia’s advanced synthetic antibody technology rapidly transforms the results of AI directed evolution into functional antibodies using high-speed DNA engineering, which can be used to create and screen more than 10 billion human synthetic antibodies in 3 – 4 weeks. Traditional non-synthetic approaches use the immune system of human or other animals to generate antibodies, resulting in antibody diversity and developability that is restricted. “BioDuro-Sundia’s synthetic antibody screening platform enables us to discover the most optimal antibodies, because we are not limited by a human or animal immune system; we can design any antibody with virtually infinite combinations of sequences that are optimized through directed evolution, and we make sure to start with a fully human antibody framework that has good development properties, meaning we have a much better likelihood of reaching the clinic—and doing so quickly.”
We are also working with several partners to extend the technology into antibody designs for immune cell therapies such as CAR-T and TCR-T, which are highly promising immuno-oncology therapeutics, but have had unexpected outcomes and strong side effects in clinical studies—largely attributed to suboptimal antibody and linker designs. The teams are applying their unique antibody discovery platform to rapidly discover and develop next-generation CAR-T therapy with improved safety and efficacy profiles.