Data Science

How Genetic Studies Can Help Us Develop Treatments for Mental Illness


  • Published1 May 2018

  • Reviewed1 May 2018



  • Author




  • Source
    BrainFacts/SfN

Worldwide, an estimated one percent of all people will develop schizophrenia. Marked by hallucinations and delusions, symptoms of this mental disorder typically begin in adolescence and early adulthood. It is a disease that does not go away and causes lifelong disability and distress to many individuals who suffer from it, and to their families.

“The lives of patients with schizophrenia are marked by multiple hospitalizations, and tragically often by homelessness and prison. And the life expectancy of people with schizophrenia, although rarely discussed, is decreased by eight to 20 years,” said Steven E. Hyman, past president of the Society for Neuroscience and director of the Stanley Center for Psychiatric Research at the Broad Institute. “We need to do more.”

Despite this critical need, mental illnesses remain difficult to understand and to treat. Here, Hyman discusses where the field has been and what lies ahead.

What treatments are currently available to treat psychiatric and behavioral disorders?

There has been more than a half-century of stagnation in the treatment of schizophrenia, bipolar disorder, and depression. Notably, there are still no pharmacologic treatments for autism spectrum disorders (ASDs) and for the cognitive impairments of schizophrenia.

Drugs that treat psychotic symptoms, such as hallucinations and delusions, were first discovered serendipitously in the 1950s. Advances since then have largely been limited to reducing side effects. There has been no advance in efficacy since the drug clozapine was discovered in the 1960s.

Why has progress been so slow?

As we know the human brain is extremely complex, but importantly, it is essentially inviolable in life, and critical aspects of human psychiatric disorders cannot be captured in animal models. For example, psychiatric disorders have been associated with anatomic abnormalities such as abnormal gray matter loss in the cerebral cortex. We have seen that in postmortem studies and noninvasive neuroimaging, but these observations do not yield molecular clues to what is causing the illnesses that could be exploited to discover new drugs.

Image of Steve Hyman

Steven E. Hyman is Director of the Stanley Center for Psychiatric Research and a member of the Broad Institute of MIT and Harvard.

Society for Neuroscience, 2017.

The high heritability of schizophrenia, bipolar disorder, ASDs, and other psychiatric disorders has long been known, but until the advent of modern genomic and computational tools, we had no way to discover the disease mechanisms. Schizophrenia and most common forms of other psychiatric disorders involve the interplay of many hundreds of genes, each contributing a small risk that quickly adds up. Scientists say these disorders are highly polygenic. As a result, psychiatric disorders are heterogeneous in that affected individuals will have different combinations of the risk-conferring versions of relevant genes.

Are we making any progress?

We are making progress. For example, genome-wide association studies (GWAS) performed on more than 80,000 people with schizophrenia have identified more than 250 regions of the genome in which genetic variation is associated with elevated risk of schizophrenia. Sequencing of the whole exomes (protein-coding regions of the genome) of more than 25,000 people with schizophrenia have identified thousands of rare variants in affected individuals, but three genes have reached exome-wide significance. Disconcertingly the rate of gene discovery shows no signs of slowing down, further evidence of the complexity of schizophrenia, but also a major challenge for neurobiologists who want to use this information to understand what goes wrong in the brain to produce this devastating illness.

How can the field speed the rate of progress?

There are limits to what we can learn with reductionist approaches like using transgenic mice. Mouse brains are different from human brains. Beyond the well-known differences in brain structure and function that reflect 90 million years of evolutionary separation in very different selective environments, there is poor conservation of regulatory regions of the genome that comprise about 90 percent of GWAS findings (across all common human diseases). In short, the genetic risk factors for schizophrenia are each small, very numerous, and peculiarly human.

We need a greater focus on human biology, ranging from human cellular models constructed with stem cell technologies to better tools to study human brain circuit activity in living humans.

Imagine, if you will, a “population in a dish” which would allow scientists to compare the cells from many individuals, ranging from those with very low genetic risk for schizophrenia to those at higher risk. If we could grow cells from hundreds of people together, thus limiting the ‘noise’ that occurs when growing each person’s cells in different dishes, we might be able to measure phenotypes (observable traits) after perturbing the cells in different ways, including with candidate drugs. New single-cell technologies permit us to re-identify the donor of each individual cell.

Another advance permits us to grow human brain organoids, tiny 3D models of human brains, in culture. These brain models give scientists the ability to observe how different risk genotypes and environmental factors influence early-stage brain development and the formation of synapses and brain circuits, which are critically affected by schizophrenia and other psychiatric disorders.

The genetic studies of psychiatric illnesses are advancing rapidly, but that ultimately helps no one if all we end up with are gene lists. It is critical that we advance the study of human biology so that we can transform genetic information into mechanistic biological knowledge and new therapeutics.


This Ask an Expert was answered by Steven E. Hyman as told to David Levine for BrainFacts.org.

BrainFacts.org welcomes all your brain-related questions.

Every month, we choose one reader question and get an answer from a top neuroscientist. Always been curious about something?