Statistical methods for analyzing immunosignatures

Abstract

BackgroundImmunosignaturing is a new peptide microarray based technology for profiling of humoral immune responses. Despite new challenges, immunosignaturing gives us the opportunity to explore new and fundamentally different research questions. In addition to classifying samples based on disease status, the complex patterns and latent factors underlying immunosignatures, which we attempt to model, may have a diverse range of applications.MethodsWe investigate the utility of a number of statistical methods to determine model performance and address challenges inherent in analyzing immunosignatures. Some of these methods include exploratory and confirmatory factor analyses, classical significance testing, structural equation and mixture modeling.ResultsWe demonstrate an ability to classify samples based on disease status and show that immunosignaturing is a very promising technology for screening and presymptomatic screening of disease. In addition, we are able to model complex patterns and latent factors underlying immunosignatures. These latent factors may serve as biomarkers for disease and may play a key role in a bioinformatic method for antibody discovery.ConclusionBased on this research, we lay out an analytic framework illustrating how immunosignatures may be useful as a general method for screening and presymptomatic screening of disease as well as antibody discovery.

Integrative Analysis of Immunological Data to Explore Chronic Immune T-Cell Activation in Successfully Treated HIV Patients

“…The simplest and best explanation model for available data maintaining the hypothesized general pathway (Fig 3) passed the goodness-of-fit tests. As estimation techniques, modeling capacities and breadth of SEM applications are expanding rapidly, beyond social sciences, to embrace multiple fields as modern immunology [41, 43, 44], endocrinology [60], cancer [61, 62] or environment [63], we would like to warn readers about SEM results that have to be interpreted carefully. Indeed, because a model fits the data well does not mean that the model reflects perfectly the reality.…”

Could immunosignatures technology enable the development of a preventative cancer vaccine?

“…Thus far, over 30 diseases have produced unique immunosignatures including multiple diseases of the pancreas [12], multiple forms of brain cancer [13] and early immunosignatures of Alzheimer's disease [14]. Stages of breast cancer were also separable using immunosignatures, including those that had a second, non-recurrent, primary tumor [15]. These immunosignatures have resulted in classification accuracies exceeding 94% [13].…”

Beta process based adaptive learning for immunosignature microarray feature identification

“…This model was refined in [17] to consider gene interdependence. A feature model based on the exploratory factor analysis approach was also considered in [18]. These models, however, did not consider possible problems with increasing number of data or complex feature relationships such as in the case of large size immunosignature microarrays.…”