By Roger Meissen | Division of Research
Few aspects are more important than control in science.
In mouse experiments, where consistent genetics mean reliable results, that control can mean the difference between medical research that can be reproduced and results that won’t hold up over time.
A new policy paper in the journal Science from the nation’s four Mutant Mouse Resource and Research Centers (MMRRC) points out substantial issues with the reported genetics of some mice they help distribute, catalogue and preserve. To address the problem, it also announces a new quality control tool to ensure better genetic accuracy.
It reveals that, in roughly half of cases, the described genetics of mice don't match the actual genetics of the animals.
“The takeaway is that half of the mice that we get in our repository are not what investigators tell us they are,” said Jim Amos-Landgraf, director of the MMRRC at MU and associate professor of Pathobiology and Integrative Biomedical Sciences. “From a rigor and reproducibility standpoint, that's problematic. If you don't know what you're working with, that can affect outcomes of research.”
Along with lead authors from the University of North Carolina and two other MMRRCs, Mizzou MMRRC examined more than 600 samples from 341 mouse strains and found inconsistencies between reported names and actual genetic makeup in half of the samples — a problem with real scientific consequences. When mislabeled mice were corrected, center staff still found that 80 percent of names failed to reflect the level of genetic variation and inbreeding.
The biomedical backbone
Lab mice genetics are a big deal for biomedical research.
Since 1901, the vast majority of Nobel Prize winners in Physiology and Medicine have depended on mouse models to make their discoveries. From cancer and heart disease to neurological disorders and countless other conditions, specially engineered mice are used to figure out how diseases work and test possible cures.
As the number of mouse models ballooned over the last century, in 1999 the National Institutes of Health funded four MMRRCs — at Mizzou, UNC, UC Davis and the Jackson Laboratory in Maine —— to catalogue, cryopreserve and distribute these models. Biomedical researchers who receive federal grants from the NIH are required to send samples of their mice models to these centers to add to this repository. Mizzou alone has more than 50,000 mice cryopreserved.
While some mice have genes edited with techniques like CRISPR, many involve inbreeding, a key to conserve and limit variation in genetics. That helps researchers focus on one question and keeps the underlying genetics from skewing results in one way or another.
“In biology, the more things you can control for, the better,” Amos-Landgraf said. “We try to limit the amount of genetic variation to be able to test a particular question, so the MMRRC has all of these inbred lines that have very unique qualities — some are resistant to heart disease, some are sensitive to heart disease, some are sensitive to cancer, some are resistant to cancer — and that genetic variation can tell us really important things about biology on a fundamental level.”
Mistaken identity
While the study points to problems, the good news is that many of the inconsistencies are often a matter of incorrect labeling instead of compromised mice.
“Just less than half of that 50 percent with incorrect genetics is just the wrong name,” Amos-Landgraf said. “They may call it C57 Black 6J, but it's actually C57 Black 6N, so very subtle differences, but significant because there are genetic differences between those that can have impacts on phenotypes.”
But some cases are more serious.
The paper describes instances where mice had unexpected genetic contamination from an entirely different strain — sometimes introduced by a simple lab error. In one example, a researcher insisted they were sending a pure black inbred mouse line, but when the MMRRC received 10 animals, one was white.
“We analyzed these mice and it turns out that there was a significant contribution of an unaccounted for genetic background.” he said. “When we went back to the investigator with that information, they acknowledged that after looking their records they had a graduate student who was responsible for that colony and had performed genetic crosses between the original line and other inbred strain we had identified, and told us that they must have sent us mice form that cross.”
The paper also flags cases where mice carried engineered DNA segments — like Cre recombinase, a tool widely used to switch genes on and off — that weren't listed in their documentation. That kind of hidden element can seriously compromise experimental results.
A tool to fix it
The solution centers on a genotyping platform called MiniMUGA (Mouse Universal Genotyping Array), developed at UNC. The technology scans tens of thousands of genetic markers and can identify the specific strain background of an animal with speed and at relatively low cost.
Crucially, Mizzou MMRRC and their partners don't test just one mouse from a submitted strain. “We use multiple samples to really determine the accuracy of what we're looking at, so that we can truly determine the level of segregation of genetic variation in that population.”
The result is a plain-language Genetic Quality Control (GQC) report that classifies each mouse strain into one of seven categories, estimates how genetically consistent the animals are and flags potential problems.
Going forward, the new testing process will be standard for every animal received by any MMRRC.
“Any animal that comes in, we're testing,” Amos-Landgraf said. “Any animal that someone then wants from us from the past, when we make it live again, we will actually do it on those as well. So once they get it in their hands, they'll know what the genetics are moving forward.”
The team also calls on university breeding programs and individual labs to adopt similar quality control practices.
The problem, Amos-Landgraf says, isn't that scientists were being careless. It's that a reliable, affordable tool to catch these errors simply didn't exist until recently. "People kind of put their head in the sand because we didn't really have a way to do that cheaply, but this tool changes that."
Now, he says, there's no reason not to know exactly what you're working with — and every reason that you should.
The MMRRC at UNC has created a series of short webinars to explain the features of the Strain GQC report.
The policy article, "Improve genetic quality control to increase rigor and reproducibility of mouse research," was published in Science May 14, 2026. Funding for MU MMRRC is provided by NIH grant U42OD010918. Read more in UNC’s release.