Systems biology strategy for analyzing innate immunity flow-cytometry data by heatmap and decision-tree analysis.
2016-01-28T12:41:28Z (GMT) by
<p>(A) Bioinformatics tool applied for single-cell data mining using heatmap computational method to preprocess flow cytometry data and to identify the innate immunity cell attributes. (B) Decision tree analysis identifies “root” (CD14<sup>+</sup>CD56<sup>+</sup>) and “secondary” (NK Granzyme A<sup>+</sup> and NK CD16<sup>+</sup>CD56<sup>-</sup>) cell attributes with higher accuracy to distinguish between non-human primates naturally infected with <i>T</i>. <i>cruzi</i> and non-infected controls. (C) Scatter distribution plots show the potential of selected biomarkers to discriminate infected from non-infected individuals. White rectangles indicate true positive (Chagas disease) and true negative (non-infected subjects) classifications. Gray rectangles indicate subjects that require the analysis of additional characteristics for accurate classification by the algorithm sequence proposed by the decision tree. (C) ROC curve analysis illustrating the cut-off points, the global accuracy (area under the curve–AUC) and performance indexes (sensitivity–Se, specificity–Sp and likelihood ratio–LR) for each selected biomarker.</p>