The AIntibody challenge: benchmarking the use of AI/ML in antibody discovery
Not withstanding extensive publicity and claims, the value of AI in antibody discovery remains unclear. The AIntibody competition was launched to benchmark real-world performance, and the potential value, of AI in antibody discovery through a blinded, prospective experimental design.
Our next High-Affinity talks features Andrew Bradbury, Chief Scientific Officer at Specifica.
In the inaugural challenge, two Specifica NGS datasets from a Generation 3 selection output were provided to 29 participating organizations that altogether submitted 527 antibody sequences responding to three competitions, for each of which the goal was to design or identify antibodies with high affinity and developability:
Given NGS datasets from three affinity maturation selection outputs with binding diversity in HCDR1/2, LCDR1/2 or LCDR3, design high affinity developable antibodies. Experimentally, these outputs are usually combined and sorted to generate affinity matured variants.
Identify the highest affinity developable antibody in each of three HCDR3 clustered selection outputs, comprising 400-3500 different sequences.
Given the full NGS selection output of challenge 2, comprising >30,000 different sequences, design high affinity out of dataset antibodies.
All sequences were expressed as full-length IgGs and experimentally tested for binding affinity using surface plasmon resonance (SPR) and developability. The affinities of the highest affinity antibodies were further validated by KinExA.
Hosted by MiLaboratories, this talk will provide a background to in vitro antibody discovery and an overview of the final competition results.