Machine Learning Framework for Predicting Vaccine Immunogenicity

flu shot

The ability to better predict how different individuals will respond to vaccination and to understand what best protects them from infection marks an important advance in developing next-generation vaccines. It facilitates the rapid design and evaluation of new and emerging vaccines. It also identifies individuals unlikely to benefit from the vaccine.

A team of researchers from the CDC, Georgia Tech, and Emory University created a general-purpose machine learning framework, called DAMIP, for discovering gene signatures that can predict vaccine immunity and efficacy.  

Using DAMIP, implemented results for yellow fever demonstrated that, for the first time, a vaccine’s ability to immunize a patient could be successfully predicted with greater than 90 percent accuracy within a week after vaccination. A gene identified by DAMIP decrypted a seven-decade-old mystery of vaccination. Results for flu vaccine demonstrated DAMIP’s applicability to both live-attenuated and inactivated vaccines. Similar results in a malaria study enabled targeted delivery to individual patients.

The project guides the rapid development of better vaccines to fight emerging infections and improve monitoring for poor responses in the elderly, infants, and those with weakened immune systems.

Importantly, the project’s work is expected to help design a universal flu vaccine.