Machine Learning Algorithm Predicts Schizophrenia Risk

By examining specific regions of the human genome, Baylor College of Medicine researchers developed a machine learning algorithm called SPLS-DA to look for epigenetic markers for schizophrenia.

The team identified epigenetic markers that differed between people diagnosed with schizophrenia and those without the condition by analyzing DNA from blood samples. With this information, the team created a model that could use predictive analytics to determine the probability of an individual having schizophrenia.

When testing the model on an independent dataset, the results revealed the model had an 80 percent accuracy rate of identifying those with the condition.

“Schizophrenia is a devastating disease that affects about 1% of the world’s population,” corresponding author Dr. Robert A. Waterland said in a press release.

“Although genetic and environmental components seem to be involved in the condition, current evidence only explains a small portion of cases, suggesting that other factors, such as epigenetic, also could be important.”

Epigenetics is a system for DNA molecular marking that tells the different cells in the body which gene to turn on or off in the cell type. Moreover, epigenetic markers can vary between different normal tissues within an individual which makes it difficult for researchers to determine if epigenetic changes contribute to diseases involving the brain, such as schizophrenia.

However, to address this issue, the team references previous work that identified a set of genomic regions in which DNA methylation differed between multiple people but was consistent across different tissues in one person. 

This region was identified as CoRSIVs, also known as correlated regions of systemic interindividual variation. The team suggested that studying CoRSIVs could uncover epigenetic causes of disease.

“Because methylation patterns in CoRSIVs are the same in all the tissues of one individual, we can analyze them in a blood sample to infer epigenetic regulation on other parts of the body that are difficult to assess, such as the brain,” said Waterland, who is also a professor of pediatrics – nutrition at the USDA/ARS Children’s Nutrition Research Center at Baylor and of molecular and human genetics.

According to the researchers, previous studies have examined methylation profiles in blood samples to identify epigenetic differences between those with schizophrenia.

Our study is innovative in various ways,” said first author Dr. Chathura J. Gunasekara, a computer scientist in the Waterland lab.

“We focused on CoRSIVs and also applied for the first time the SPLS-DA machine learning algorithm to analyze DNA methylation. As a scientist interested in applying machine learning to medicine, our findings are very exciting. They not only suggest the possibility of predicting risk of schizophrenia early in life, but also outline a new approach that may be applicable to other diseases.”

The current study expands on previous research by considering potential confounding factors. For example, factors such as smoking and taking antipsychotic medications can impact methylation patterns in the blood. These factors are common in schizophrenia patients.

“Here, we took various approaches to evaluate whether the methylation patterns we detected at CoRSIVs were affected by medication use and smoking. We were able to rule that out,” Waterland said.

“This, together with the fact that DNA methylation at CoRSIVs is established very early in life, indicates that the epigenetic differences we identified between schizophrenia patients and healthy individuals were there before the disease was diagnosed, suggesting they may contribute to the condition.”

By using this machine learning approach, researchers achieve much stronger epigenetic signals associated with schizophrenia than done before, according to the team.

“We consider our study a proof of principle that focusing on CoRSIVs makes epigenetic epidemiology possible,” Waterland said.