Given the existing flow cytometry software, why did we spend so much time, money, and effort to create AutoGate? What does this software do that the others do not?
The answer boils down to a simple concept – automation that you control, as the preview screen below illustrates. In no time, AutoGate lets you preview all the clusters in all the samples in whatever experiment you are doing, which helps you decide at any point which sample you want to look at next.
To the degree and method you choose, you can turbocharge your work, asking AutoGate to do do all your compensation and gating for you.
In addition to AutoGate's original semi-automatic gating which we describe above, AutoGate in the last 3 years has acquired 5 fully automatic gating methods. Two methods are unsupervised: Exhaustive Projection Pursuit (EPP) and Uniform Manifold Approximation and Projection (UMAP). Three methods are supervised: multilayer perceptron (MLP) neural network based on MATLAB fitcnet; MLP based on Python TensorFlow and UMAP supervised templates.
Below is a comparison of the results with 7 publications based on average mass+distance similarity (QFMatch) and by overlap (F-measure)
Link to Publication
T cells, dendritic
B cells, macrophages
EPP is designed to find subsets based on phenotyping markers and scatter parameters.
AutoGate does multi-dimensional scaling (MDS) of subset medians with Matlab's built-in function cmdscale.
AutoGate produces a visual dendogram to express HiD readiness