On October 14, 2026, Dallas-based Colossal Biosciences announced GenoMorph, an AI platform designed to identify gene-editing targets for dodo revival. Using hundreds of samples from chickens, quail, ducks, and pigeons across six embryonic-development stages and three facial regions, the system identified approximately 2,300 candidate regulatory DNA elements: more than 650 associated with beak size and nearly 1,700 with beak thickness and breadth. The candidates are entering validation studies, with confirmed elements potentially edited into Nicobar pigeon primordial germ cells, because the Nicobar pigeon is the dodo’s closest living relative.[1]
GenoMorph’s primary significance is methodological: it extracts likely regulatory instructions from living relatives when the extinct organism’s genomic evidence is limited. Regulatory DNA controls when and how strongly genes are expressed, so the approach could help reconstruct phenotypes without possessing a complete, directly interpretable genome. For the Lunar Ark, this supports a preservation strategy based on layered biological archives—genome sequences, developmental data, living reference organisms, cell lines, and computational models—rather than DNA storage alone. It also exposes major risks: candidate associations are not yet proven causal, and edited organisms require extensive validation for developmental stability, fertility, welfare, ecological containment, and unintended mutations.[1]
The Ark team should track GenoMorph’s validation results, editing accuracy, germline transmission, and performance in Nicobar pigeon cells; preserve comparable multi-species developmental datasets; and develop auditable AI pipelines that record training data, model versions, uncertainty, edits, and experimental outcomes. Priority research should include cryopreserved primordial germ cells, avian embryo culture, regulatory-genome annotation, and simulation environments for testing restoration plans before release. Colossal’s reported use of AI to compress projects that might otherwise take more than 50 years indicates that restoration systems should treat biological reconstruction as an iterative, version-controlled engineering discipline while retaining independent biosafety review.[3]