Gallax Achieves Success in 8 out of 9 Cases in AI Antibody Design Comparison, Outperforming Competitors with 4-5 Cases

Comparison with Nabla Bio and Chai Discovery's Published Studies Using Common Targets

"Proving AI's Competitiveness in Designing Antibodies from Scratch"

Gallax, an artificial intelligence (AI) drug development company, announced on May 19 that its proprietary AI protein design platform, "Gallax Design," outperformed competitors in a global comparison of de novo antibody design technologies.

Galaxy company logo image. Galaxy

De novo antibody design is a technology in which AI designs entirely new antibodies capable of binding to specific proteins, rather than identifying existing antibodies. It is considered a next-generation technology that can accelerate the discovery of new drug candidates.

This study was conducted based on nine protein targets that could be commonly validated in recently published global de novo antibody design studies. The comparison included JAM-2 from Nabla Bio in the United States and Chai-2 from Chai Discovery.

Gallax explained that it designed 50 antibody candidates for each target and experimentally verified that it secured antibodies binding to the actual targets in eight out of the nine cases. Under the same criteria, Nabla Bio secured binding antibodies for five targets, and Chai Discovery for four targets.

The company emphasized that these results not only represent a higher number of successes, but also demonstrate the capability to consistently design antibodies across a diverse range of targets. Until now, each company had announced results based on different conditions and targets, making direct comparisons difficult. This study, however, compared performance using common targets.

However, the company noted that producing antibodies that bind to a target does not immediately lead to the development of a therapeutic drug. For a candidate to become an actual new drug, further validation of safety, efficacy, and manufacturability is required. Nevertheless, the company assessed that these results are meaningful in that they can significantly accelerate the early discovery of candidates.

Gallax also previously announced last year that its AI-designed antibody candidates achieved a binding success rate of over 30% across eight different target sites.

Tae-yong Park, Vice President of Gallax, said, "AI antibody design technology is advancing beyond simple proof-of-concept toward reliably generating desired antibodies for a wide range of targets," adding, "This will greatly accelerate the pace of early candidate discovery."

In a comparison based on 9 common targets, Galaxy Design succeeded in obtaining bispecific antibodies for 8 targets, showing the highest target-specific success rate compared to JAM-2 with 5 targets and Chai-2 with 4 targets. Galaxy

This content was produced with the assistance of AI translation services.