Using clustering algorithms, Schier and his colleagues grouped fish into 60,000 distinct behavioral profiles, depending on various constraints. "When you turn off the light, how often are they active? When they are inactive, how long? That's what we observe in the fish," says Schier. "You can measure many different parameters, and that allows you to profile different drugs."

Anti-inflammatories, such as cytokines, nonsteroidal anti-inflammatory drugs, and cyclosporine, had a surprising effect. Normally, these drugs induce sleep when taken to combat infection such as the flu. However, Schier found that when given to normal, healthy zebrafish, these compounds, or immunomodulators, made fish more active during the day.

"In disease, immunomodulators have been implicated in sleep," says Schier. "We propose that maybe there's some baseline function for these immunomodulators during normal sleep and wake cycles."

Such findings could help researchers identify new molecular players involved in sleep and wakefulness. Irina Zhdanova, associate professor of anatomy and neurobiology at Boston University Medical School, studies the physiological mechanisms of circadian rhythms and sleep in zebrafish. Zhdanova says there are many sleep-related drugs on the market with substantial side effects; these effects might be avoided with better screening tools.

"The huge scope of drugs tested [by Schier's group] shows that zebrafish-based tests can be effectively used to at least prescreen multiple classes of existing drugs and new candidate substances," says Zhdanova. "[That is] certainly very helpful."

In the future, Schier says, zebrafish could also be used as a model for testing drugs for human psychiatric diseases like schizophrenia and autism. The idea is to identify genes associated with the human disease, and try to engineer the same genetic defect in zebrafish. Researchers could then look for certain behavioral changes as a result, such as a fish's sensitivity to touch, or its reaction to visual cues.

"Hopefully there would be a connection between the gene affected, and change in behavior, and one would try to correct the change in behavior by adding particular drugs," says Schier. "That's a bit science fiction at the moment, but it is possible."