Tim Cernak's University of Michigan lab uses AI driven protein modeling and automated synthesis to design species specific drugs for Gila monsters, bald eagles, loggerhead sea turtles, and trees, in a young field he calls conservation chemistry.
Tim Cernak spent roughly two decades inside a Big Pharma drug discovery pipeline before a wildlife case reframed the work. The specifics of that moment, which animal and which compound, sit at the center of why he now runs a small University of Michigan lab that completes about 1,500 reactions a day, almost none of them aimed at humans (MIT Technology Review).
Cernak calls the discipline "conservation chemistry," a label he coined in 2018 to describe a narrow, still-young practice: designing new drugs for animal patients when no approved treatment exists, or when the standard of care is itself toxic to the patient. It is not a recognized subdiscipline yet, and Cernak's group is one of a handful trying to define what the work can and cannot do (MIT Technology Review).
The method is concrete enough to picture. Cernak's team uses AI-driven protein structure prediction and benchtop synthesis robots to screen candidate molecules virtually, then push the most promising through automated reactions at a pace no human chemist could match. The candidates are tested on animal cell lines and, where collaborators allow, on the actual patients: Gila monsters, loggerhead sea turtles, hemlock trees under pressure from invasive beetles, and bald eagles infected with H5N1 (MIT Technology Review).
The reason the work exists, as MIT Technology Review frames it, is a documented chain of contamination. Human and veterinary pharmaceuticals reach wildlife through water, soil, and food: drug residues, livestock treatments, and accidental exposures. Cernak's patient list is itself a catalog of that chain. H5N1 in bald eagles is tied in part to industrial poultry practices, putting a raptor at the end of a contamination pathway it did not choose. Conservation chemistry sits downstream of these problems, building a capability to design compounds for the species most exposed and least served (MIT Technology Review).
The honest limit is upstream of all of this. Conservation chemistry addresses one slice of pharmaceutical contamination: the species for which no safe drug exists, and for which designing a new one is feasible. It does not address the source problem of drug residues entering water and soil in the first place, including manufacturing effluent, prescribing practices, and agricultural runoff. Critics in adjacent fields argue that source reduction would save more animals per dollar than any numbers of new molecules. Cernak's work is a complement to that effort, not a substitute (MIT Technology Review).
The next 12 to 24 months will be the first real test. Cernak's lab has compounds moving toward animal trials for several of the species on its roster, and the discipline's first widely cited success would be a wildlife drug approved on species-specific safety data rather than a human drug repurposed and tolerated. A partial outcome, where compounds work in cell lines but fail in whole animals, would still teach the field where its bottlenecks sit, and would clarify whether conservation chemistry scales beyond a single lab (MIT Technology Review).