On October 28, 2026, Dallas-based Colossal Biosciences reported using its machine-learning platform GenoMorph to identify approximately 2,300 DNA regions associated with the dodo’s beak and facial structure. GenoMorph compares embryonic genetic data from chickens, quail, ducks, and pigeons with genomes and skull measurements from more than 70 bird species. The work relies on a single high-quality dodo genome; the identified regions remain laboratory-test candidates, not confirmed trait controls. Colossal is targeting the Nicobar pigeon as the dodo’s living genetic relative, and no date for a living dodo has been announced.[1]
The principal technical lesson is that AI can infer candidate genotype-to-phenotype relationships when reference genomes are sparse, but restoration remains constrained by biological validation, delivery systems, developmental compatibility, and ecological uncertainty. Colossal’s reported progress with lipid nanoparticles for avian gene-editing delivery and Nicobar pigeon cells is relevant to maintaining viable cell lines, embryos, and genetically diverse founder populations in a lunar repository. De-extinction also exposes existential-risk controls: engineered organisms could create unforeseen welfare, ecological, or pathogen hazards if released without containment and multigenerational testing.[1]
The Ark team should monitor GenoMorph’s validation results, avian gene-editing delivery methods, Nicobar pigeon germline experiments, and the number and diversity of preserved avian genomes, cells, embryos, and reproductive protocols. The Ark should integrate a trait-linked genomic archive that stores raw sequence data, uncertainty estimates, developmental reference data, phenotype measurements, and reproducible software—not merely edited genomes. Priority should go to preservation of living relatives, cryogenic cell banks, and ecosystem-function datasets, with restoration plans gated by containment, genetic diversity, disease screening, and ecological impact assessments.[1]