HMN 2026: How AI can provide new insights into why Ebola outcomes differ between patients

Using AI to provide new insights into why Ebola outcomes differ between patients
Dr. Isabel Garcia Dorival conducting field research in Guinea during the 2015 Ebola outbreak while a post-doctoral associate at the University of Liverpool. She is currently a principal investigator and research scientist at the Instituto de Salud Carlos III (ISCIII) in Madrid, Spain. Credit: University of Liverpool

There is currently a serious outbreak in the Democratic Republic of the Congo of the severe and often fatal illness. Symptoms include fever, vomiting, diarrhea, severe dehydration and hemorrhage. Rapid clinical deterioration is common. University of Liverpool researchers have led two major new studies into Ebola virus disease (EVD) and factors that influence and predict patient outcomes.

During outbreaks, health care workers currently rely on viral load—the amount of virus present in your body—to predict which patients are most at risk of severe disease or death. However, viral load alone does not reliably explain why some patients with similar levels of virus survive while others do not.

Two new collaborative studies published in The Journal of Infectious Diseases, led by Professor Julian Hiscox, have now provided important molecular clues that could improve patient triage and deepen understanding of how age and sex influence disease severity. Both studies used blood samples from hospitalized patients from the 2013–2016 outbreak in West Africa, where his lab deployed under the European Mobile Laboratory.

In the first study, researchers used machine-learning approaches to examine what changes in the host immune response were associated with survival. They identified several biomarkers that differed significantly between survivors and fatal cases. When these host markers were combined with viral load, the accuracy of predicting clinical outcome increased substantially. These results demonstrated the potential for further diagnostics to support clinical decision-making in future outbreaks.

The second study explored how age and sex influence the host immune response. People who survived the infection tended to have a less severe and more controlled immune response, although the specific ways their bodies managed this differed depending on sex and age. One key finding was that genes involved in lymphocyte differentiation decreased with age in fatal cases but increased with age in survivors. These differences suggest that age and sex should be considered when developing future treatments.

Professor Hiscox stated, “These collaborative studies help explain why Ebola affects people so differently and highlight the importance of understanding the host response, not just the virus itself. Our findings will support better clinical management and guide the development of more effective diagnostics and treatments.”

“The FDA is proud to support innovative research that advances our understanding of high-consequence infectious diseases like Ebola,” said FDA Chief Scientist Dr. Steven Kozlowski, M.D. “Global collaboration is foundational to enabling us to better understand Ebola virus disease, helping lay a foundation for more precise diagnostics, improved patient management, and stronger public health preparedness.”

Publication details

Jocelyn G Pérez et al, Identification of host gene transcripts by machine learning and their application to predict outcome in Ebola virus disease, The Journal of Infectious Diseases (2026). DOI: 10.1093/infdis/jiag308

Xiaofeng Dong et al, Molecular phenotypes of sex and age specific differences relating to outcome in patients with Ebola virus disease, The Journal of Infectious Diseases (2026). DOI: 10.1093/infdis/jiag305

Journal information:
Journal of Infectious Diseases


Clinical categories

Infectious diseases

This content is provided for information purposes only.