The idea of scientists trying to grow brain tissue in a dish conjures up all sorts of scary mental pictures (cue the horror-movie music). But the reality of the research is quite far from that sci-fi vision—and always will be, say researchers in the field. In fact, a leader in this area of research, Arnold Kriegstein of the University of California, San Francisco, says the reality does not measure up to what some scientists make it out to be.

In a paper published on January 29 in Nature, Kriegstein and his colleagues identified which genes were active in 235,000 cells extracted from 37 different organoids and compared them with 189,000 cells from normally developing brains. The organoids—at times called “mini brains,” to the chagrin of some scientists—are not a fully accurate representation of normal developmental processes, according to the study.

Brain organoids are made from stem cells that are transformed from one cell type to the another until they end up as neurons or other mature cells. But according to the Nature paper, they do not always fully complete this developmental process. Instead the organoids tend to end up with cells that have not fully transformed into new cell types—and they do not re-create the normal brain’s organizational structure. Psychiatric and neurodevelopmental conditions—including schizophrenia and autism, respectively—and neurodegenerative diseases such as Alzheimer’s are generally specific to particular cell types and circuits.

Many of the organoid cells showed signs of metabolic stress, the study demonstrated. When the team transplanted organoid cells into mice, their identity became “crisper,” and they acted more like normal cells, Kriegstein says. This result suggests that the culture conditions under which such cells are grown does not match those of a normally developing brain, he adds. “Cellular stress is reversible,” Kriegstein says. “If we can reverse it, we’re likely to see the identity of cells improve significantly at the same time.”

Brain organoids are getting better at recapitulating the activities of small clusters of neurons, says Kriegstein, who is a professor of neurology and director of the Eli & Edythe Broad Center for Regeneration Medicine and Stem Cell Research at U.C.S.F. Scientists often make organoids from the cells of people with different medical conditions to better understand those conditions. But some scientists may have gone too far in making claims about insights they have derived from patient-specific brain organoids. “I’d be cautious about that,” Kriegstein says. “Some of those changes might reflect the abnormal gene expression of the cells and not actually reflect a true disease feature. So that’s a problem for scientists to address.”

A small ball of cells grown in a dish may be able to re-create some aspects of parts of the brain, but it is not intended to represent the entire brain and its complexity, several researchers have asserted. These organoids are no more sentient than brain tissue removed from a patient during an operation, one scientist has said.

Of course, models are never perfect. Although animal models have led to fundamental insights into brain development, researchers have sought out organoids, or organs-in-a-dish, precisely because of the limitations of extrapolating biological insights from another species to humans. Alzheimer’s has been cured hundreds of times in mice but never in us, for instance.

“That said, the current models are already very useful in addressing some fundamental questions in human brain development,” says Hongjun Song, a professor of neuroscience at the Perelman School of Medicine at the University of Pennsylvania, who was not involved in the new research. Using brain organoids, he adds, the Zika virus was recently shown to attack neural stem cells, causing a response that could explain why some babies exposed to Zika in utero develop unusually small brains.

Michael Nestor, a stem cell expert, who did not participate in the new study, says his own organoids are very helpful for identifying unusual activity in brain cells grown from people with autism. And he notes that they will eventually be useful for screening potential drugs.

Even though the models will always be a simplification, the organoid work remains crucial, says Paola Arlotta, chair of the department of stem cell and regenerative biology at Harvard University, who was also not involved in the Nature study. Neuropsychiatric pathologies and neurodevelopmental conditions are generally the result of a large number of genetic changes, which are too complex to be modeled in rodents, she says.

Sergiu Pasca, another leader in the field, says that the cellular stress encountered by Kriegstein and his team might actually be useful in some conditions, helping to create in a dish the kinds of conditions that lead to diseases of neurodegeneration, for instance. “What I consider the most exciting feature remains our ability to derive neural cells and glial cells in vitro, understanding their intrinsic program of maturation in a dish,” says Pasca, an assistant professor at Stanford University, who was not part of the new paper.

The ability to improve cell quality when exposed to the environment of the mouse brain suggests that it may be possible to overcome some of the current limitations, Arlotta says. There is not yet a single protocol for making brain organoids in a lab, which may be for the best at this early stage of the field. Eventually, she says, scientists will optimize and standardize the conditions in which these cells are grown.

Arlotta, who is also the Golub Family Professor of Stem Cell and Regenerative Biology at Harvard, published a study last year in Nature showing that she and her colleagues can—over a six-month period—make organoids capable of reliably including a diversity of cell types that are appropriate for the human cerebral cortex. She says it is crucial for organoid work to be done within an ethical framework. Arlotta is part of a federally funded team of bioethicists and scientists working together to ensure that such studies proceed ethically. The scientists educate the bioethicists on the state of the research, she says, and the ethicists inform the scientists about the implications of their work.

Nestor feels so strongly about the importance of linking science, policy and public awareness around stem cell research that he has put his own laboratory at the Hussman Institute for Autism on hold to accept a year-long science-and-technology-policy fellowship with the American Association for the Advancement of Science.  He says he took the post to make sure the public and policy makers understand what they need to know about organoids and other cutting-edge science and to learn how to communicate about science with them.

One thing all of the scientists interviewed for this article agree on is that these brain organoids are not actual mini brains, and no one is trying to build a brain in a dish. Even as researchers learn to make more cell types and grow them in more realistic conditions, they will never be able to replicate the brain’s structure and complexity, Kriegstein says. “The exquisite organization of a normal brain is critical to its function,” he adds. Brains are “still the most complicated structure that nature has ever created.”