AI may be able to predict how well a vaccine will work for you before you even get the shot. #
-
Date:
-
August 21, 2026
-
Source:
-
Arizona State University
-
Summary:
-
AI may be able to predict how strongly someone will respond to a vaccine before they receive it. By analyzing antibody patterns in more than 4,000 people, researchers found signs of “immune readiness” that helped distinguish strong responders from weak ones. Surprisingly, even some healthy people responded poorly, while some immunosuppressed people mounted strong responses.
-
Share: Vaccines prevent serious illness for many people, but the immune protection they produce can differ substantially from one person to another. New research led by Arizona State University offers clues about what may be behind those differences.
The immune system may show signs of how strongly it will react even before vaccination. Researchers at ASU and collaborating institutions examined blood samples from more than 4,000 people, measuring antibodies that recognized 185 antigens. Those immune targets included common viruses and bacteria, along with targets connected to autoimmune diseases.
Artificial intelligence was then used to search for patterns in blood samples taken before and after COVID-19 vaccination. The analysis uncovered antibody signatures that could help separate people who produced strong vaccine responses from those whose responses were weaker.
The findings could eventually contribute to vaccination strategies tailored more closely to an individual's immune system.
"What our study found is that certain biomarkers, when analyzed with AI, can predict who is likely to respond well to a vaccine, even before they receive it. This suggests that some people may be more immune-ready than others," says Joshua LaBaer, who led the study.
LaBaer is executive director of the Biodesign Institute at ASU and director of the Virginia G. Piper Center for Personalized Diagnostics. The project also involved ASU researchers and collaborators from medical and research institutions around the United States.
The study appears in the current issue of the journal Cell Press Blue.
Blood Antibodies May Reveal Vaccine Readiness
Scientists typically evaluate vaccine response after vaccination by measuring whether the immune system generated antibodies against the intended target. In this study, the researchers approached the problem from the opposite direction. They wanted to know whether immune patterns already present in the blood could reveal how someone would respond before receiving a vaccine.
Many factors can influence vaccine response, including age, sex, genetics, previous illnesses and underlying health conditions. People with conditions that compromise the immune system are often more likely to produce weaker responses. However, vaccine outcomes can still vary widely among people who fall into the same general health categories.
The researchers used one of the first approaches to examine a broad antibody "fingerprint" present before vaccination as a measure of immune readiness. While some other prediction strategies depend on genetic testing, this method analyzes antibody patterns in blood, potentially making it easier to translate into clinical practice.
Health Status Alone Does Not Predict Response
To investigate whether these antibody fingerprints could signal vaccine readiness, the team measured immune responses to 185 antigens. The targets included SARS-CoV-2, the virus responsible for COVID-19, as well as other widespread viruses and bacteria and targets associated with autoimmune diseases.
Altogether, the researchers examined 8,687 samples from 4,089 participants. The group included healthy volunteers as well as people with diseases or treatments associated with immune suppression, including HIV, multiple myeloma, solid organ malignancy, autoimmune disease, inflammatory bowel disease and solid organ transplantation.
Several immunosuppressed groups were more likely to show reduced responses to COVID-19 vaccination. Yet simply placing someone into an immunosuppressed or healthy category did not reliably predict the outcome.
Some participants with suppressed immune systems still developed strong responses. At the same time, about 5% to 6% of healthy participants showed weak vaccine responses.
"Sentinel" Antibodies Signal Immune Readiness
Certain antibodies that were already present before vaccination stood out in the analysis. Higher levels of antibodies targeting common microbes, including Staphylococcus aureus, RSV and human respirovirus 3, were associated with stronger responses to COVID-19 vaccines.
The researchers call these "sentinel" antibodies because they may serve as indicators of a person's underlying immune readiness. These antibodies are not necessarily acting directly against the vaccine target. Instead, their presence may provide information about how prepared the antibody-producing portion of the immune system is to mount a response.
The team also investigated whether the complete antibody fingerprint could provide more predictive information than a small number of individual biomarkers. A deep learning model examined patterns across the entire antibody panel, combining numerous measurements to build a broader picture of each participant's immune state.
AI Searches Millions of Immune Signals
The results demonstrate one potential advantage of using AI in biomedical research. Machine learning systems can search millions of biological data points for subtle relationships that may be difficult to detect using conventional approaches.
In this case, the findings suggest that understanding vaccine readiness may require looking at the immune system as an interconnected whole instead of concentrating on one antibody or one disease.
The research also demonstrates the potential of newer technologies capable of measuring many antibody responses simultaneously. Rather than testing whether a person has antibodies against a single pathogen, researchers can examine a much broader immune landscape shaped by previous exposure to viruses, bacteria and other immune targets.
Toward More Personalized Vaccination
If the findings are confirmed in future studies and extended to additional vaccines, the approach could have uses well beyond COVID-19. Profiling sentinel antibodies might eventually support vaccine research, vaccine development and medical care for people who are especially vulnerable to weak immune responses. Doctors could potentially use this type of information to identify people who might benefit from additional vaccine doses, more careful follow-up or other protective strategies. It could also give scientists a clearer understanding of why vaccination produces powerful immune responses in some people but weaker ones in others.
Ultimately, the research points toward a future in which vaccination decisions could be informed by an individual's own level of immune readiness.
Story Source:
Materials provided by Arizona State University. Note: Content may be edited for style and length.
Journal Reference:
- Lusheng Song, Jin G. Park, Ji Qiu, Vel Murugan, Seyedmasoud Mousavi, Ching-Wen Hou, D. Mitchell Magee, Zhongxue Chen, Yunro Chung, Stacy Williams, Deborah Adam, Bihong Guo, Candyce McDaniel, Toria Trendler, Gabe Rice, Kelly Connard, Bradley Bobbett, Michael Ritchie, Izamar Garcia, Guillermo Sebastian Trivino Soto, Giovanna Caruth, Eleni Katergaris, Kristen Seifert, Veronica Boyle, Heewook Lee, Damodara Rao Mendu, Viviana Simon, Komal Srivastava, Bharat Thyagarajan, Mike Ryan, Elena Kowalsky, Patrick Breads, David Fetterer, Troy Kemp, Ann Khalsa, Lora Nordstrom, Michael D. White, Raquel Salgado, Chad M. VanDenBerg, Natacha Montalvo, Craig Woods, Rajat Walia, David Carpentieri, Peter Arden, Christine L. Kuryla, Anay Patel, Lauren Stiene, Imke Folkerts, Stefani N. Thomas, Anne Davidson, Thomas Hickey, Nancy Roche, Ligia Pinto, Norman Kleiman, Carlos Cordon-Cardo, Amy B. Karger, Peter Gregersen, Joshua LaBaer. Pre-vaccine sentinel antibodies predict blunted vaccine responses.Cell Press Blue, 2026; 100088 DOI:10.1016/j.cpblue.2026.100088
Cite This Page:
ScienceDaily. Retrieved August 21, 2026 from www.sciencedaily.com