Figure 1 | Scientific Reports

Figure 1

From: Risk assessment of latent tuberculosis infection through a multiplexed cytokine biosensor assay and machine learning feature selection

Figure 1

Workflow for LTBI supernatant sample analysis. (A) Subject PBMCs are stimulated under multiple on- and off-target conditions. (B) Samples are analyzed using the Genalyte Matchbox system, which uses plug-and-play chip and device interfaces to measure cytokine concentrations quickly and reproducibly in a multiplexed assay format. The resonance shift output is recorded and converted to concentrations based on individual cytokine calibrations in the sample matrix. (C) Random Forest bioinformatics determine what clinical features are essential for categorical distinctions and predictive accuracy based on variable importance metrics, with statistical data reduction methods employed to identify biomarker signatures most highly correlated with given clinical determinants.

Back to article page