In October 2026, Colossal Biosciences announced GenoMorph, an AI platform that identified approximately 2,300 candidate regulatory DNA elements associated with dodo beak and facial morphology. The set includes 653 elements linked to beak size and 1,698 linked to beak thickness and breadth. The candidates were derived from comparative genomics, developmental biology, and high-resolution 3D morphology across living birds; they remain hypotheses requiring laboratory validation, not confirmed causal genes.[1][2]
The principal technical advance is prioritisation under severe data loss. Rather than recreating a complete dodo genome, Colossal plans to engineer selected traits into the Nicobar pigeon, the dodo’s closest living relative, using regulatory elements inferred from chickens, quail, ducks, pigeons, and other bird species. The reported use of lipid nanoparticles for genomic editing within an egg could also improve delivery of edits during early avian development.[1][7] For the Lunar Ark, this establishes a relevant model for reconstructing biological functions from fragmented genomes, phenotype records, and surrogate organisms, while highlighting the danger of confusing AI-ranked correlations with validated causal mechanisms.
The Ark team should monitor GenoMorph’s validation results, embryo-editing outcomes, off-target effects, developmental abnormalities, fertility, and multi-generation stability. It should also research an equivalent restoration pipeline: maintain redundant genome and phenotype archives, preserve regulatory DNA and developmental datasets, model candidate edits with uncertainty scores, and define experimental validation gates before attempting restoration. Priority should go to organisms supporting food systems, pollination, nutrient cycling, and human health—not symbolic species alone.[2][4]