Making the data speak

Mapping Genes for Personalized Care

Ma works in a subfield known as genomics, which studies genes and how they function inside cells. New next-generation sequencing technologies allow scientists to examine tissues in unprecedented detail, revealing which genes are active in individual cells and generating vast amounts of molecular data. Ma develops statistical and computational methods that make sense of all that information.

She aims to understand why diseases such as cancer or Alzheimer’s progress differently from person to person. Ultimately, she hopes her work will help uncover the biological mechanisms driving those differences so that the knowledge can be used to improve diagnosis, sharpen prediction and, one day, guide more personalized treatment.

“When you collect the data, it doesn’t look like a picture of tissue,” Ma said. “It’s just measurements showing which genes are expressed in each cell. We use statistical methods to reconstruct those numbers into a clear picture of what’s happening inside the tissue.”

For example, in cancer research, scientists want to understand how different cells are arranged inside a tumor. Researchers look for regions dominated by cancer cells and regions that contain immune cells trying to attack the tumor, because that balance can shape how aggressive the tumor is, how it progresses and how it responds to treatment. Those patterns are not visible just by looking at raw data. They have to be uncovered through the type of careful analysis Ma and her team at Brown conduct. 

Once that spatial distribution is identified within a single tissue sample, the next step is comparison with thousands of other samples. Measuring how those patterns differ from person to person helps pinpoint what is driving those differences at the molecular level and can help explain why a disease develops or progresses differently in different patients. Over time, those insights can improve doctors’ ability to anticipate how a patient’s illness is likely to behave.

“That math can help the doctors make better decisions or make better treatment plans,” Ma said.

This type of work is the focus of a recently funded federal project Ma was awarded by the National Institutes of Health. The goal is to link these cellular patterns to real-world outcomes, such as how advanced a cancer is, or how well a patient may respond to a particular therapy.

Advancing Risk Prediction

In recent years, Ma has also become a leader in what’s known as polygenic risk scores, which estimate a person’s inherited risk for certain diseases based on their genetic profile. Ma has helped develop new ways of building these scores and has published widely on how they can be used responsibly and effectively.

In one project, published in The American Journal of Human Genetics, Ma and her collaborators created an online resource that compiles polygenic risk scores for dozens of common health-related traits, from cholesterol levels to cancer risk. The platform allows researchers around the world to access, compare and test these scores in large health databases.

“This allows researchers to better understand genetic contributions to disease risk and through that may help identify individuals who could benefit from earlier monitoring or preventive strategies,” Ma said.

Prioritizing Data Privacy

Still, as genetic data becomes more abundant, so do concerns about how that information is protected. For instance, genetic information can reveal deeply personal details about a person’s health and ancestry, making it especially sensitive. 

Through a National Science Foundation grant, Ma is helping develop statistical and computational tools that allow hospitals and research centers to collaborate without exposing sensitive patient information. Her approach can also help reduce bias and improve how well predictive models work across different communities by enabling institutions to learn from larger and more diverse patient populations.

“There is a long way to go, but we are hoping to make it work because…this can lead to more robust discoveries and models that generalize better across populations while protecting data privacy,” Ma said. It also reminds Ma that the entire subfield she’s in—mapping where genes and molecules are located inside tissues—is still a relatively new area of study.

Even so, she is encouraged by the rapid technological advances of the past two decades and our growing ability to analyze information at a scale that once seemed impossible. This, Ma says, applies not just to genomics but to all of biostatistics, including at Brown. It is all part of an effort, as Ma put it, to “make the data speak.”