reoptimize_strategy: reoptimize_strategy

In ebecht/hypergate: Machine Learning of Hyperrectangular Gating Strategies for High-Dimensional Cytometry

reoptimize_strategy

Description

Optimize a gating strategy given a manual selection of channels

Usage

reoptimize_strategy(
  gate,
  channels_subset,
  xp,
  gate_vector,
  level,
  beta = 1,
  verbose = FALSE
)

Arguments

gate	A return from hypergate
channels_subset	Character vector identifying the channels that will be retained (others are ignored). The form is e.g. c("CD4_min","CD8_max")
xp	Expression matrix as in the hypergate call
gate_vector	Categorical vector as in the hypergate call
level	Level of gate_vector identifying the population of interest
beta	Yield / purity trade-off
verbose	Whether to print information about optimization status

Examples

data(Samusik_01_subset)
xp=Samusik_01_subset$xp_src[,Samusik_01_subset$regular_channels]
gate_vector=Samusik_01_subset$labels
hg=hypergate(xp=xp,gate_vector=gate_vector,level=23,delta_add=0)
contribs=channels_contributions(gate=hg,xp=xp,gate_vector=gate_vector,level=23,beta=1)
significant_channels=names(contribs)[contribs>=0.01]
hg_reoptimized=reoptimize_strategy(gate=hg,channels_subset=significant_channels,xp,gate_vector,23)

ebecht/hypergate documentation built on Feb. 4, 2024, 3:29 p.m.