Dataset Information

Multiple network-constrained regressions expand insights into influenza vaccination responses.

ABSTRACT:

Motivation

Systems immunology leverages recent technological advancements that enable broad profiling of the immune system to better understand the response to infection and vaccination, as well as the dysregulation that occurs in disease. An increasingly common approach to gain insights from these large-scale profiling experiments involves the application of statistical learning methods to predict disease states or the immune response to perturbations. However, the goal of many systems studies is not to maximize accuracy, but rather to gain biological insights. The predictors identified using current approaches can be biologically uninterpretable or present only one of many equally predictive models, leading to a narrow understanding of the underlying biology.

Results

Here we show that incorporating prior biological knowledge within a logistic modeling framework by using network-level constraints on transcriptional profiling data significantly improves interpretability. Moreover, incorporating different types of biological knowledge produces models that highlight distinct aspects of the underlying biology, while maintaining predictive accuracy. We propose a new framework, Logistic Multiple Network-constrained Regression (LogMiNeR), and apply it to understand the mechanisms underlying differential responses to influenza vaccination. Although standard logistic regression approaches were predictive, they were minimally interpretable. Incorporating prior knowledge using LogMiNeR led to models that were equally predictive yet highly interpretable. In this context, B cell-specific genes and mTOR signaling were associated with an effective vaccination response in young adults. Overall, our results demonstrate a new paradigm for analyzing high-dimensional immune profiling data in which multiple networks encoding prior knowledge are incorporated to improve model interpretability.

Availability and implementation

The R source code described in this article is publicly available at https://bitbucket.org/kleinstein/logminer .

Contact

steven.kleinstein@yale.edu or stefan.avey@yale.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

SUBMITTER: Avey S

PROVIDER: S-EPMC5870750 | biostudies-literature | 2017 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Multiple network-constrained regressions expand insights into influenza vaccination responses.

Avey Stefan S Mohanty Subhasis S Wilson Jean J Zapata Heidi H Joshi Samit R SR Siconolfi Barbara B Tsang Sui S Shaw Albert C AC Kleinstein Steven H SH

Bioinformatics (Oxford, England) 20170701 14

<h4>Motivation</h4>Systems immunology leverages recent technological advancements that enable broad profiling of the immune system to better understand the response to infection and vaccination, as well as the dysregulation that occurs in disease. An increasingly common approach to gain insights from these large-scale profiling experiments involves the application of statistical learning methods to predict disease states or the immune response to perturbations. However, the goal of many systems ...[more]

PMID: 28881994

Similar Datasets

Project description:The strength and breadth of an individual's antibody repertoire is an important predictor of their response to influenza infection or vaccination. Although progress has been made in understanding qualitatively how repeated exposures shape the antibody mediated immune response, quantitative understanding remains limited. We developed a set of mathematical models describing short-term antibody kinetics following influenza infection or vaccination and fit them to haemagglutination inhibition (HI) titres from 5 groups of ferrets which were exposed to different combinations of trivalent inactivated influenza vaccine (TIV with or without adjuvant), A/H3N2 priming inoculation and post-vaccination A/H1N1 inoculation. We fit models with various immunological mechanisms that have been empirically observed but have not previously been included in mathematical models of antibody landscapes, including: titre ceiling effects, antigenic seniority and exposure-type specific cross reactivity. Based on the parameter estimates of the best supported models, we describe a number of key immunological features. We found quantifiable differences in the degree of homologous and cross-reactive antibody boosting elicited by different exposure types. Infection and adjuvanted vaccination generally resulted in strong, broadly reactive responses whereas unadjuvanted vaccination resulted in a weak, narrow response. We found that the order of exposure mattered: priming with A/H3N2 improved subsequent vaccine response, and the second dose of adjuvanted vaccination resulted in substantially greater antibody boosting than the first. Either antigenic seniority or a titre ceiling effect were included in the two best fitting models, suggesting a role for a mechanism describing diminishing antibody boosting with repeated exposures. Although there was considerable uncertainty in our estimates of antibody waning parameters, our results suggest that both short and long term waning were present and would be identifiable with a larger set of experiments. These results highlight the potential use of repeat exposure animal models in revealing short-term, strain-specific immune dynamics of influenza.

Dataset Information

Multiple network-constrained regressions expand insights into influenza vaccination responses.

Motivation

Results

Availability and implementation

Contact

Supplementary information

Publications

Multiple network-constrained regressions expand insights into influenza vaccination responses.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure