In DillonHammill/CytoGate: openCyto Plugin Gating Functions

library(knitr)
opts_chunk$set(out.extra='style="display:block; margin: auto"', fig.align="center", message = FALSE, warning = FALSE)

Overview of CytoGate Features

Introduction

The Cytogate package provides an interactive interface for drawing gates to analyse flow cytometry data through the DrawGate function. The package also contains a plugin function gate_draw to incorporate these interactive features into the openCyto gating pipeline. Here we aim to demonstrate the main gating features of CytoGate and how these features can be used with openCyto.

1. Adding CytoGate to openCyto as a Plugin.

The features of CytoGate can be easily incorporated into openCyto as a plugin. To do this we will need to supply the name of the gating function gate_draw and give the gating method a name for use within openCyto, here we will use DrawGate as the gating method name. To check whether the function has been successfully added we use listgtMethods() to return a list of supported gating methods

library(openCyto)
library(CytoGate)

registerPlugins(fun = gate_draw, methodName = "DrawGate")
listgtMethods()

2. Loading in Data for Analysis

To demonstrate the main features of CytoGate we will use a data set supplied by CytoGate called Activation. This data set contains eight .fcs files used to determine the activation state of OT-I and OT-II transgenic T cells. We will add these samples to a GatingSet called gs for subsequent gating using the openCyto framework.

fs <- Activation

gs <- GatingSet(fs)

3. Preparing Data for Analysis

Once the data has been added to the GatingSet we perform necessary compensation using the attached spillover matrix and perform logicle transformation on all fluorescent channels.

spill <- fs[[1]]@description$SPILL
gs <- compensate(gs, spill)

channels <- colnames(fs)
trans.channels <- channels[!channels %in% c("FSC-A","FSC-H","FSC-W","SSC-A","SSC-H","SSC-W","Time")]

trans <- estimateLogicle(gs[[1]], trans.channels)
gs <- transform(gs, trans)

4. Supported Gate Types & Expected Inputs

The DrawGate gating method will automatically open a new interactive plotting window where the user can specify the coordinates of gates manually. The type of gate to constructed is supplied to the GatingTemplate as the gate_type argument - this will be demonstarted later. Currently the supported gate types and expected inputs are as follows:

"polygon" constructs a polygon gate from n selected points where n must be greater than 3. To seal the polygon the user must right click and select "stop". This is the default gate_type used by DrawGate. n>3
"rectangle" constructs a rectangle gate parallel to the x and y axes from 2 selected points along the diagonal of the rectangle. n=2
"interval" constructs a gate based on user defined lower and upper bounds. 1D and 2D interval gates are supported, For 1D gates a density distribution is plotted and 2 points selected indicating the lower and upper bounds of the gate. 2D interval gates require an additional argument axis used to specifiy whether the "x" or "y" axis should be gated. n=2
"threshold" constructs a gate that selects all events above a selected lower bound. 1D and 2D threshold gates are supported. For 2D threshold gates a rectangle gate is constructed which selects events above the selcted x and y co-ordinates. n=1
"ellipse" constructs an ellipse gate from 4 selcted points indicating the limits of the gate in 2 dimensons. n=4
"quadrant" constructs 4 rectangle gates based on selection of a single point indicating the center of the crosshairs. n=1

Multi-gates are supported for all gate types except "threshold" and "quadrant" through an additional argument N defining the number of gates to be constructed. This multi-gate feature will be demonstrated later.

5. Polygon Gates (n)

Below is an example of the entry in the GatingTemplate. When using DrawGate as an openCyto plugin it is recommended that pooled using the collapseDataForGating and groupBy arguments to ensure that the selected gates encompass the population in all samples. Additionally, to speed up the gating process user should also supply the subSample argument to restrict the data to a subset of the pooled data for plotting. Notice how the gate_type is supplied as a gating_args using 'gatetype'.

template <- add_pop(
  gs, alias = "Cells", parent = "root", pop = "+", dims = "FSC-A,SSC-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='polygon'", collapseDataForGating = TRUE, groupBy = 8
)

6. Rectangle Gates (n=2)

template <- add_pop(
  gs, alias = "PE+", parent = "root", pop = "+", dims = "FSC-A,SSC-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='rectangle'", collapseDataForGating = TRUE, groupBy = 2
)

6. Interval Gates

1-Dimensional Interval Gates (n=2)

template <- add_pop(
  gs, alias = "PE+", parent = "root", pop = "+", dims = "PE-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='interval'", collapseDataForGating = TRUE, groupBy = 2
)

2-Dimensional Interval Gates - Axis Argument (n=2)

template <- add_pop(
  gs, alias = "PE+", parent = "root", pop = "+", dims = "APC-Cy7-A,PE-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='interval',axis='y'", collapseDataForGating = TRUE, groupBy = 2
)

7. Threshold Gates

1-Dimensional Threshold Gates (n=1)

template <- add_pop(
  gs, alias = "PE+", parent = "root", pop = "+", dims = "PE-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='threshold'", collapseDataForGating = TRUE, groupBy = 2
)

2-Dimensional Threshold Gates (n=1)

template <- add_pop(
  gs, alias = "PE+FITC+", parent = "root", pop = "+", dims = "Alexa Fluor 488-A,PE-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='threshold'", collapseDataForGating = TRUE, groupBy = 2
)

8. Ellipsoid Gates (n=4)

template <- add_pop(
  gs, alias = "PE+FITC+", parent = "root", pop = "+", dims = "Alexa Fluor 488-A,PE-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='ellipse'", collapseDataForGating = TRUE, groupBy = 2
)

9. Quadrant Gates (n=1)

Quadrant gates returns 4 gates and therefore 4 populations names must be supplied as the alias argument and pop should be set to "*" indicating that multiple gates will be returned.

template <- add_pop(
  gs, alias = "DN,FITC+,AF700+FITC+,AF700+", parent = "root", pop = "*", dims = "Alexa Fluor 488-A,Alexa Fluor 700-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='quadrant'", collapseDataForGating = TRUE, groupBy = 2
)

10. Multi-Gates

Here we will demonstrate the multi-gate feature of DrawGate for gate_type = "ppolygon" only although this feature is supported for all other gate types except threshold and quadrant. Simply supply the number of gates to construct as the argument N - for example to construct 2 polygon gates set N=2, supply two population names to the alias column and set pop="*" to return multuiple populations.

template <- add_pop(
  gs, alias = "PE+,FITC+", parent = "root", pop = "*", dims = "Alexa Fluor 488-A,PE-A", gating_method = "DrawGate",
  gating_args = "subSample=25000,gate_type='polygon',N=2", collapseDataForGating = TRUE, groupBy = 2
)