Cardiff, Wales: Antiverse, a biotechnology company developing a computational antibody drug discovery platform, today announced a partnership with a top 20 pharma company with core competencies in the life science fields of health care and nutrition. The collaboration has been successful in identifying antibody candidates for a target of interest, with greater diversity (2.3x) and accuracy compared to alternative bioinformatics pipeline selection methods.
Using Antiverse’s proprietary AI-Augmented Drug Discovery platform, 248 sequences were identified, of which over 230 were antibody binders. Five of these antibody candidates were found to be cross-reactive.
Antiverse is combining machine learning and phage display techniques to model antibody-antigen interactions. The current version of the platform uses next generation sequencing and AI to provide diverse antibody candidates for any given target. The technology is being developed to enable the development of drugs for “difficult” targets associated with cancer, heart, and lung diseases. The platform will overcome limitations commonly associated with traditional screening including overcoming amplification bias which leads to diversity loss and selection bias which results in the loss of unique antibodies.
Murat Tunaboylu, Chief Executive Officer of Antiverse, said: “We are pleased to collaborate with a top 20 pharma company on this discovery project. Our AI-powered antibody drug discovery platform successfully identified more binders with better diversity to state-of-the-art high-throughput colony picking and bioinformatics pipeline selection. Researchers can find more diverse antibodies by unlocking the full potential of their library/discovery process with our AI-augmented discovery platform and through our fee-for-service offering, we hope to partner with other biotech and pharma companies on projects to identify other challenging targets.”
For more information about Antiverse’s AI-powered antibody drug discovery platform, please visit: https://www.antiverse.io/#what