Colossal Biosciences announced Wednesday the launch of GenoMorph, an artificial intelligence platform designed to solve a fundamental problem in de-extinction: identifying which genes to edit when you have almost no genetic data.
The bottleneck is real. Reviving an extinct species typically requires comparing hundreds of genomes to pinpoint trait-specific genetic differences. That works for living species. It fails for the dead. The dodo project, for instance, has access to exactly one complete dodo genome. The woolly mammoth effort, by contrast, has 100 genomes—a luxury afforded by better DNA preservation in frozen soil.
GenoMorph solves this by learning from the living. The platform ingests developmental data from multiple bird species—chickens, quail, ducks, pigeons—combined with genetic information and 3D skull measurements from more than 70 bird species. It then identifies which DNA regions activate during beak and facial structure development and predicts the genetic switches that may have produced the dodo's distinctive morphology.
The result: approximately 2,300 editable targets for the dodo's cranial-facial traits, ready for lab testing. CEO Ben Lamm called it a "treasure map for the dodo genome." In practical terms, it converts a vast search space into a workable list.
Colossal also built a delivery mechanism—lipid nanoparticles that ferry gene-editing tools into developing bird embryos—so researchers can observe changes without waiting years for birds to reach sexual maturity. Compress the feedback loop, compress the R&D timeline.
Peter Diamandis has launched a new fund to back startups using this platform, signaling that venture capital sees a business model here. De-extinction tech, if it works, creates IP around species-specific genetic edits and the AI systems that identify them. The economics depend on speed: reducing the years-to-viability window is the entire value proposition.
