Influenza virus has a devastating societal impact, causing up to 650,000 deaths every year worldwide. Specifically, vulnerable are children and elderly. It is estimated that 1 in 1000 children and elderly every year are hospitalized due to influenza infection. Vaccination can prevent influenza-like illnesses, and thus lower the risk of the virus outbreak. However, currently available vaccines do not always provide protection, even among otherwise-healthy people, leading to serious pandemics. Development of better vaccines depends on our understanding why current vaccines work in some individuals, while fail in others.

Recent advances in the computational biology and the development of novel machine learning algorithms make it possible to extract knowledge and identify patterns in an unbiased manner from large clinical datasets. Application of machine learning algorithms to clinical datasets can reveal immune cell populations and genes that mediate stronger antibody responses to influenza vaccines. Specifically, it is of the utmost importance to include single-cell analysis at the protein level, such as mass cytometry combined with multiple high-dimensional biological measurements, since these have power to reveal heterogeneity of the immune system and capture interindividual variability.

To identify cellular biomarkers that can predict antibody responses to inactivated influenza vaccine, we developed a tool to automate machine learning process - SIMON - Sequential Iterative Modelling “OverNight”. SIMON is a fully automated binary classifier that utilizes 128 different machine learning algorithms to select relevant and informative features. Such an iterative approach facilitated data reduction and improved overall performance, ultimately increasing the predictive accuracy. Preprint available @BioRxiv.