Fig. 3

Higher dimensional data is difficult to visualize concisely. a With highly-multiplexed single-cell data, displaying the breakdown of different functional groups being secreted by a sample can increase by up to a factor of 2× with the addition of x cytokines. When this analysis is performed across a set of donors or stimulation conditions, effectively highlighting the key secretion differences is challenging. In this standard bar graph visualization of functional groups secreted by CD4+ CAR-T cells of four donors, it is cumbersome to see which are the major functional groups being secreted by each donor, and what are the biggest fold differences across donors. b-c Reducing the dimensionality of the dataset is a different approach to more effective and understandable visualizations. PCA (principal component analysis) uses an orthogonal transformation to convert the original dataset into a set of linearly uncorrelated principal components, where the number of components is smaller than the number of original variables. The transformation is defined in such a way that the first principal component has the largest possible variance (accounting for as much variability as possible within the dataset), followed by the second component, and so on. While reducing the dimensionality to two principal components may still result in some loss of information, the benefit is that the transformed data points can then be visualized on a two-dimensional scatterplot. In this figure, PCA is applied to the 4-donor CD4+ CAR-T secretion dataset. Each cell’s secretions (signal intensity of each cytokine) are log transformed prior to dimensionality reduction. b is color-coded by donor, while c is color-coded by some of the individual cytokines. The combination of these graphs reveals some information, such as the low overall polyfunctionality of donor 2, and the high Granzyme B + MIP-1a + polyfunctionality of Donor 4. However, more detailed information about upregulated and/or distinct polyfunctional subsets is less clear