Mucking about in FlowJo is fun, especially when you can get immediate visual feedback from changing a transformation setting. For the below, I've chosen to plot only in Pseudocolor and vary only one setting at a time in FlowJo Mac version 9.3.1. For a series of lovely movies that manipulate all three settings at once, check out the Daily Dongle's Width Basis Titration.
In short, reducing the width basis value compresses the linear space around zero. See here for settings at -1, -10, -100, and -1000. It results in more or less plotspace devoted to the Negative and Dim populations. I find that customizing this setting to be most useful for investigating dim events.
For the positive decades, the setting reflects the first decade that is displayed after the linear space ends. It feels like the plot zooms-in on the intermediately bright populations. See here for settings as applied at 3, 3.5, 4, 4.5, and 5.
When it comes to the additional negative decades, settings result in more and more space devoted in plotspace for negative events. See here for 0, 0.5, 1, 1.5. I don't find altering them very useful.
I found that using a positive control for douplets helps figure out the best gating method for doublet discrimination. Use this Doublet Discrimination on the FACSAria PDF to see plotting of Area, Height and Width for chicken erythrocyte nuclei from BD's DNA QC Particle Kit. I'd be happy to hear your comments.
With thanks to the Emory Vaccine Center's flow cytometry core and Dr. John Altman.
In a collaborative effort with FlowJo, Steve Perfetto, and Ryan Brinkman's group, I'm collecting statistics from manual gating on bead files. The pupose is to define a reference range of statistics that adequately describes the variation from one manual analyst to another. The files in the analysis set are completely artificial, and eliminate any experience bias that may occur if real-world samples were provided to a variety of individuals possessing different levels of exposure to real population gating. The context for this collection is development of an evaluation method for automated gating algorithm. More on automated cluster detection and the related Q and B platform in development can be found on the FlowJo and FlowDx websites.
Members of the Brinkman lab are well-known in the open-source community of Bioconductor/R for their flow cytometry packages, and the exercise is expected to have significant benefit to developers on any platform. Would you like to get involved in cluster development too? Or you can download the files and instructions here. Send an email to my FlowJo account: suzanne at treestar dot com and I can walk you through the analysis. The gist of it is in the pic above.