NEST-Score

This repository includes the R package NEST-Score (NEighbourhood Sample homogeneiTy-Score) as introduced in “Reconstitution of Human Brain Cell Diversity in Organoids via Four Protocols” (Naas et al. 2024, bioRxiv).

It is compatible with a typical Seurat(Hao et al. 2024) workflow to analyse single-cell RNA-sequencing (scRNA-seq) data.

Installation

You can install the development version of NEST-Score R Package with:

# install.packages("devtools")
devtools::install_github("jn-goe/NESTScore")

Example

This is an exemplary workflow using NEST-Score on the PBMC 3k dataset provided by SeuratData (see the respective GitHub repository). For details on the scRNA-seq object pre-processing see the corresponding Seurat tutorial:

library(Seurat)
library(SeuratData) 
library(ggplot2)
library(NESTscore)

data('pbmc3k') 
obj <- UpdateSeuratObject(pbmc3k)

# quality control
obj[["percent.mt"]] <- PercentageFeatureSet(obj, pattern = "^MT-")
obj <- subset(obj, subset = nFeature_RNA > 200 & nFeature_RNA < 2500 & percent.mt < 5)

# pre-processing 
obj <- NormalizeData(obj)
obj <- FindVariableFeatures(obj, nfeatures = 2000)
obj <- ScaleData(obj)
obj <- RunPCA(obj)
obj <- RunUMAP(obj, reduction = "pca", dims = 1:10)

DimPlot(obj, group.by = "seurat_annotations")

Let’s create an artificial sample assignment sample_random for every cell in the dataset, which has non-uniform global frequencies and shows a near-perfect mixedness of samples across the whole dataset:

obj$sample_random <- sample(c(rep(1,3),rep(2,3),3,4), dim(obj)[1], replace = T)
obj$sample_random <- paste0("sample ", obj$sample_random)
table(obj$sample_random)
#> 
#> sample 1 sample 2 sample 3 sample 4 
#>     1009      987      319      323

DimPlot(obj, group.by = "sample_random")

To not only visually evaluate how well the samples mix, we can compute the cell-wise NEST-Score as follows:

NESTres <- NESTscore(obj, group_by = "sample_random", k_nn = 30, ndims = 50)
FeaturePlot(NESTres$seuratobj, feature = "NESTscore_sample_random", order = T) +
  scale_color_viridis_c(limits = NESTres$NESTscore_limits)

We can observe uniformly high NEST-Scores since all samples are represented in all cell neighborhoods (k_nn = 30 nearest neighbors in the ndims = 50 dimensional Principal Component space).

For comparison, let’s create a second assignment, where two cell clusters are consisting of only one sample, respectively:

obj$sample_grouped <- sample(1:2, dim(obj)[1], replace = T)
obj$sample_grouped[obj$seurat_annotations %in% c("B")] <- 3
obj$sample_grouped[obj$seurat_annotations %in% c("CD14+ Mono",
                                                "FCGR3A+ Mono",
                                                "DC")] <- 4

obj$sample_grouped <- paste0("sample ", obj$sample_grouped)
DimPlot(obj, group.by = "sample_grouped")

Again, we can compute the NEST-Score as follows:

NESTres <- NESTscore(obj, group_by = "sample_grouped", k_nn = 30, ndims = 50)
FeaturePlot(NESTres$seuratobj, feature = "NESTscore_sample_grouped", order = T) +
  scale_color_viridis_c(limits = NESTres$NESTscore_limits)

For this assignment we can see, that cells of samples 3 and 4 have low NEST-Scores since they do not mix with any cells of samples 1 and 2. On the other hand, cells of sample 1 and 2 mix and hence have higher NEST-Scores.

The NEST-Score function also provides pairwise evaluations of how well two considered samples mix on average:

NESTres <- NESTscore(obj, group_by = "sample_random", k_nn = 30, ndims = 50, 
                     return_pairwise_eval = T, show_heatmap = T)

NESTres <- NESTscore(obj, group_by = "sample_grouped", k_nn = 30, ndims = 50, 
                     return_pairwise_eval = T, show_heatmap = T)

References

Hao, Yuhan, Tim Stuart, Madeline H. Kowalski, Saket Choudhary, Paul Hoffman, Austin Hartman, Avi Srivastava, et al. 2024. “Dictionary Learning for Integrative, Multimodal and Scalable Single-Cell Analysis.” Nature Biotechnology 42 (2): 293–304. https://doi.org/10.1038/s41587-023-01767-y.

Naas, Julia, Meritxell Balmana, Laurenz Holcik, Maria Novatchkova, Lina Dobnikar, Thomas Krausgruber, Sabrina Ladstatter, et al. 2024. “Reconstitution of Human Brain Cell Diversity in Organoids via Four Protocols.” bioRxiv, 2024.11.15.623576. https://doi.org/10.1101/2024.11.15.623576.