In sistm/CytOpt-R: Optimal Transport for Gating Transfer in Cytometry Data with Domain Adaptation

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

CytOpT

CytOpT uses regularized optimal transport to directly estimate the different cell population proportions from a biological sample characterized with flow cytometry measurements.

Overview

CytOpT is an R package that provides a new algorithm relying regularized optimal transport to directly estimate the different cell population proportions from a biological sample characterized with flow cytometry measurements. Algorithm is based on the regularized Wasserstein metric to compare cytometry measurements from different samples, thus accounting for possible mis-alignment of a given cell population across sample (due to technical variability from the technology of measurements).

The main function of the package is CytOpT().

The methods implemented in this package are detailed in the following article:

Paul Freulon, Jérémie Bigot, Boris P. Hejblum. CytOpT: Optimal Transport with Domain Adaptation for Interpreting Flow Cytometry data. Annals of Applied Statistics, 17(2), 1086-1104. doi:10.1214/22-AOAS1660 https://doi.org/10.1214/22-AOAS1660 https://arxiv.org/abs/2006.09003

Installation

You can install and load CytOpT into R from CRAN with the following commands:

install.packages("CytOpT")
library(CytOpT)

Alternatively, you can install the development version of CytOpT like so:

remotes::install_github("sistm/CytOpT-R")
library(CytOpT)

Example

This is a basic example of CytOpt usage:

Data import

library(CytOpT)
# Load source Data
data("HIPC_Stanford")

# Define the true proportions in the target data set
gold_standard_manual_prop <- c(table(HIPC_Stanford_1369_1A_labels)/length(HIPC_Stanford_1369_1A_labels))

Proportion estimations using optimal transport and minmax swapping procedures

# Run CytOpt and compare the two optimization methods
res <- CytOpT(X_s = HIPC_Stanford_1228_1A, X_t = HIPC_Stanford_1369_1A, 
              Lab_source = HIPC_Stanford_1228_1A_labels,
              theta_true = gold_standard_manual_prop,
              eps = 0.0001, lbd = 0.0001, n_iter = 10000, n_stoc=10,
              step_grad = 10, step = 5, power = 0.99, 
              method='both', monitoring=TRUE)

summary(res)

plot(res)

Bland_Altman(res$proportions)

sistm/CytOpt-R documentation built on June 1, 2025, 6:12 p.m.