Figure 2
From: A novel data fusion method for the effective analysis of multiple panels of flow cytometry data

Fusion model after variable selection. The left panel (a) shows the obese individuals (blue crosses) and the lean controls (red circles). If the predicted value was above the threshold, the individuals were classified as obese. The three panels on the right (b–d) present the weights in the model of, respectively, the B cell, T cell and innate datasets. Areas coloured blue belong to cells more abundant in obese individuals, while those in red belong to cells more common in lean (control) individuals. The blue contours show where, on average, 80% of the cells of the obese individuals lie, while the red contours indicate 80% of the cells in the lean individuals.