Biomarkers and causal key drivers of phenotypic heterogeneity in peanut allergy

点击翻译按钮获取中文摘要

Biomarkers and causal key drivers of phenotypic heterogeneity in peanut allergy PROJECT 3 SUMMARY / ABSTRACT Unexpected allergic reactions to peanut are the most common cause of fatal food-related anaphylaxis. There is currently no method to predict reaction thresholds for subjects with peanut allergy. Given two peanut allergic subjects with similar clinical profiles, one may react with anaphylaxis to minute exposure while the other may react with hives to larger amounts. The same individuals may have disparate responses to desensitization that cannot be predicted. This heterogeneity in reaction threshold, reaction severity, and desensitization success is crippling to peanut allergic individuals, whose lives are impaired by anxiety that small exposures could lead to anaphylaxis at any time. Additionally, providers cannot offer early guidance whether resource-intensive desensitization efforts will succeed. With peanut allergy now affecting 2-5% of US schoolchildren, these areas of uncertainty stress the need to identify biomarkers of reaction threshold, reaction severity, and desensitization potential. Based on our demonstrated work in biomarker development and integrative genomics, we hypothesize that biomarkers and causal key drivers of phenotypic heterogeneity in peanut allergy can be identified through integrated network-based examination of peripheral blood transcriptomes, epitope-binding, and clinical parameters from peanut allergic subjects. We will study peanut allergic subjects undergoing oral food challenges and desensitization to pursue three specific aims that address unmet needs in peanut allergy care and knowledge. In Aim 1, we will identify a resting-state peripheral blood biomarker of exquisitely sensitive (low threshold) peanut allergy by RNA-sequence profiling baseline peripheral blood from low and high threshold peanut allergic children, differential gene expression analysis, machine learning, and weighted gene coexpression network analysis. The predictive biomarker of reaction threshold identified will be prospectively validated. In Aim 2, we will identify causal key drivers of peanut allergy severity. We will use baseline and post-challenge RNAseq profiles, expression quantitative trait loci, epitope-binding scores, and clinical variables to build the first probabilistic causal network specific to food allergy. We will apply key driver analysis to this network to identify genes, epitopes, and clinical variables that causally modulate peanut reaction severity. This information-rich, data-driven network will be shared with other investigators seeking to elucidate mechanisms underlying peanut allergy. In Aim 3, we will identify an early-appearing peripheral blood biomarker of peanut desensitization potential using samples obtained from subjects during desensitization, leukocyte deconvolution, machine learning, and weighted gene coexpression network analysis. The results from this project will directly address unmet needs in the management of peanut allergy by identifying peripheral blood biomarkers that predict reaction threshold and desensitization potential in peanut allergic individuals. The project will also further our mechanistic understanding of peanut allergy severity. Although we focus on peanut allergy, we expect that many findings will be applicable to other food allergies.