Abstract:Despite decades of research on the influenza virus, we still lack a predictive understanding of how vaccination reshapes each person's antibody response, which impedes efforts to design better vaccines. Here, we combined fifteen prior H3N2 influenza vaccine studies from 1997-2021, collectively containing 20,000 data points, and demonstrate that a person's pre-vaccination antibody titers predicts their post-vaccination response. In addition to hemagglutination inhibition (HAI) titers against the vaccine strain, the most predictive pre-vaccination feature is the HAI against historical influenza variants, with smaller predictive power derived from age, sex, BMI, vaccine dose, the date of vaccination, or geographic location. The resulting model predicted future responses even when the vaccine composition changed or a different inactivated vaccine formulation was used. A pre-vaccination feature ‒ the time between peak HAI across recent variants ‒ distinguished large versus small post-vaccination responses with 73% accuracy. As a further test, four vaccine studies were conducted in 2022-2023 spanning two geographic locations and three influenza vaccine types. These datasets formed a blinded prediction challenge, where the computational team only received the pre-vaccination data yet predicted the post-vaccination responses with 2.2-fold error, comparable to the 2-fold intrinsic error of the experimental assay. This approach paves the way to better utilize current influenza vaccines, especially for individuals who exhibit the weakest responses.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: how to use an individual's antibody titer (especially the antibody titer against multiple H3N2 influenza variants) before influenza vaccination to predict their antibody response after vaccination. This research aims to analyze a large amount of historical data and establish a model that can accurately predict an individual's response to the influenza vaccine, so as to better understand which factors affect the effectiveness of the vaccine and why the same vaccine produces different effects in different people. Specifically, the goals of the paper include: 1. **Combining data from multiple studies**: Integrate data from 15 H3N2 influenza vaccine studies between 1997 and 2021. These data contain approximately 20,000 data points to show that an individual's pre - vaccination antibody titer can predict their post - vaccination response. 2. **Identifying the most predictive features**: Besides the hemagglutination - inhibition (HAI) titer against the vaccine strain, the feature that is most predictive of the post - vaccination response is the HAI titer against historical influenza variants. Other factors with less predictive power include age, gender, body mass index (BMI), vaccine dose, vaccination date, or geographical location. 3. **Establishing a prediction model**: Develop an algorithm to use an individual's pre - vaccination HAI titer to predict their peak HAI response 3 - 4 weeks after vaccination, with an emphasis on the H3N2 subtype. This model uses only pre - vaccination measurements as input and does not need to "peek" at the initial vaccine response. 4. **Verifying the accuracy of the model**: Verify the predictive ability of the model through four new vaccine studies conducted in 2022 - 2023 (covering two geographical regions and three influenza vaccine types). The calculation team received only pre - vaccination data but was able to predict the post - vaccination response with an error of only 2.2 - fold, which is comparable to the inherent error (2 - fold) determined experimentally. 5. **Exploring the relationships among predictive variables**: The study found that the most critical variables for predicting the post - vaccination response are an individual's pre - vaccination HAI titer against the vaccine strain, pre - vaccination HAI titer against other influenza variants, whether they received an inactivated vaccine or a live - attenuated vaccine, and their influenza exposure history. In addition, the time interval of the HAI peak of the most recent variant (\(\Delta \text{Peak}\)) can also distinguish between a larger post - vaccination response and a smaller response with an accuracy of 73%. Through these goals, this study hopes to provide a scientific basis for the design of more effective influenza vaccines and better protection strategies for individuals with a weak response to existing vaccines.

Leveraging Pre-Vaccination Antibody Titers across Multiple Influenza H3N2 Variants to Forecast the Post-Vaccination Response

Quantifying the impact of pre-vaccination titre and vaccination history on influenza vaccine immunogenicity

Leveraging baseline transcriptional features and information from single-cell data to power the prediction of influenza vaccine response

Using Interpretable Machine Learning to Massively Increase the Number of Antibody-Virus Interactions Across Studies

Timely vaccine strain selection and genomic surveillance improves evolutionary forecast accuracy of seasonal influenza A/H3N2

Repeated vaccination with homologous influenza hemagglutinin broadens human antibody responses to unmatched flu viruses

High-throughput sequencing-based neutralization assay reveals how repeated vaccinations impact titers to recent human H1N1 influenza strains

Predicting Influenza H3N2 Vaccine Efficacy From Evolution of the Dominant Epitope

Prediction of Influenza B Vaccine Effectiveness from Sequence Data

Predicting the Mutating Distribution at Antigenic Sites of the Influenza Virus.

H3N2 Influenza Infection Elicits More Cross-Reactive and Less Clonally Expanded Anti-Hemagglutinin Antibodies Than Influenza Vaccination.

Measures of population immunity can predict the dominant clade of influenza A (H3N2) in the 2017-2018 season and reveal age-associated differences in susceptibility and antibody-binding specificity

Multicohort analysis reveals baseline transcriptional predictors of influenza vaccination responses

A sequence-based machine learning model for predicting antigenic distance for H3N2 influenza virus

The Potential Benefits of Delaying Seasonal Influenza Vaccine Selections for the Northern Hemisphere: A Retrospective Modeling Study in the United States

Multiplexed Antibody Sequencing and Profiling of the Human Hemagglutinin-specific Memory B Cell Response following Influenza Vaccination

Improved resolution of influenza vaccination responses with high-throughput live virus microneutralisation

Relationship between haemagglutination-inhibiting antibody titres and clinical protection against influenza: development and application of a bayesian random-effects model

Imprinting of Repeated Influenza A/H3 Exposures on Antibody Quantity and Antibody Quality: Implications for Seasonal Vaccine Strain Selection and Vaccine Performance

Applying Unique Molecular Identifiers in Next Generation Sequencing Reveals a Constrained Viral Quasispecies Evolution under Cross-Reactive Antibody Pressure Targeting Long Alpha Helix of Hemagglutinin

Computational Identification of Antigenicity-Associated Sites in the Hemagglutinin Protein of A/H1N1 Seasonal Influenza Virus.