Positivity Thresholding

This is the web-supplement to the manuscript entitled Optimizing the F-measure to set positivity thresholds. We include the FCS data used to generate the results in the manuscript, code to implement the methods described therein and a step-by-step example of how to obtain a positivity threshold using the provided data and code.

Contents

• An example application
  • Getting started
  • Read the files into python using py-fcm.
  • Define the channels
  • Find the positivity threshold
  • Get the percentages, counts, and indices—then plot
• FCS files and other relevant data
  • FCS files
  • Code
  • MIFlowCyt
• Algorithm implementation
  • List of functions
  • Get positivity threshold
  • Calculate fscore
  • Get cytokine positive events
• A multi-center example—EQAPOL
  • Methods
    • EQAPOL - Methods
      • Building a statistical model of the data
      • Assigning events to clusters based on their statistical properties
      • Classifying clusters into basic lymphocyte cell subsets
      • Defining positivity thresholds for cytokine positive events
      • Literature cited
  • Results
    • Basic subsets – CD3
    • Basic subsets – CD4
    • Basic subsets – CD4
    • Cytokine subsets – CD4
    • Cytokine subsets – CD8
    • Summary
• Automated clustering methods
  • Figure 4
    • Methods