ThresholdR uses Gaussian Mixture Models concept to identify the noise population in each surface marker across cells in CITE-seq experiments. Once the noise distribution is identified, it calculates the upper threshold of the noise component to separate expressing and non-expressing cells for each surface marker.
Oliaeimotlagh, M., Kumar, S., Taraskin, A., Suthahar, S. S. A., Suryawanshi, V., Chiang, A. W., & Ley, K. (2025). Automated denoising of CITE-seq data with ThresholdR. Cell Reports Methods, 5(7). DOI: 10.1016/j.crmeth.2025.101088
#1. Install required packages:
required_packages <- c("mixtools", "mclust", "foreach", "ggplot2", "AdaptGauss")
# Check and install missing packages
for (pkg in required_packages) {
if (!requireNamespace(pkg, quietly = TRUE)) {
install.packages(pkg)
}
}
#2. Install ThresholdR:
devtools::install_github('MDMotlagh/ThresholdR')| Folder | Description | Details |
|---|---|---|
| data | bmcite dataset | bmcite is available through SeuratData R package. |
| vignettes | The example codes | Vignette/bmcite_example.Rmd and Vignette/bmcite_example.html |
| output | Fitting plots and threshold values | Vignette/02Plots/ and Vignette/03AllThresholds.csv |