🧠 Food for thought

Implications, context, and why it matters.

The funding is backed by specific performance metrics

• Galux’s AI platform, GaluxDesign, ties its pitch to measured outcomes that likely drew investor interest.
• It achieved a 31.5% overall hit rate by generating 50 AI-designed antibodies per epitope (the part of a target protein an antibody binds to), with 10.5% reaching therapeutically meaningful affinity 1.
• Those results produced drug candidates such as an antibody for the cancer target PD-L1 (a protein some tumors use to evade immune attack) with a 9.0 picomolar binding affinity (KD=9.0 pM) plus developability comparable to the commercial therapeutic antibody atezolizumab 2.
• The system also designs antibodies for targets without an experimentally resolved structure, such as ALK7, and creates antibodies that separate a mutant from a wild-type protein based on one amino acid change, such as EGFR(S468R) vs. wild-type EGFR 2.

AI is enabling a new biopharma business model

• The funding signals a wider move from probabilistic drug discovery that screens massive libraries to rational, AI-driven drug design 1.
• Large pharmaceutical companies increasingly rely on partners instead of keeping every step of R&D in-house.
• In Galux’s partnership with Celltrion, Galux handles AI-driven antibody engineering plus early-stage candidate validation, while Celltrion manages non-clinical and clinical development through commercialization 3.
• This setup supports more complex treatments such as multi-specific antibodies (engineered antibodies that can bind more than one target) that hit several disease-related targets with a single molecule, which the companies describe as difficult to develop with traditional random screening approaches 3.