scDEED (single-cell dubious embeddings detector): a statistical method for detecting dubious non-linear embeddings

• This package is used to determine the reliability of non-linear dimension reduction embeddings. It provides functions to detect dubious cells and trustworthy cells in tSNE and UMAP embeddings. Furthermore, by minimizing the number of dubious cells, functions in this package find the best perplexity parameter of tSNE and the best n.neighbors/min.dist parameter of UMAP.

• Choose the suitable dimension for PCA (num_pc)

• Input count matrix should contain cells as columns and genes as rows

Installation

You can install the released version of scDEED from GitHub with:

library(devtools)
devtools::install_github("JSB-UCLA/scDEED")

Example

This is a basic example showing how to find the best parameter.If users use our example input data and the default parameter list(s), users can get the result in about 2mins We use an example input data which is generated by randomly sampled 10000 cells from Hydra data as a demo:

suppressPackageStartupMessages(library(scDEED))
data(input_counts)

Choose the suitable dimension for PCA (num_pc)

chooseK(input_counts)

ChooseK plot:

Example for umap

umap_example <- scDEED(input_counts , num_pc = 16, use_method = "umap",visualization = TRUE)

head(umap_example$`number of dubious cells corresponding to pair of n.neighbors and min.dist list`)

	n.neighbors	min.dist	number of dubious cells
1	5	0.1	42
2	6	0.1	54
3	7	0.1	39
4	8	0.1	76
5	9	0.1	29
6	10	0.1	43

umap_example$`best pair of n.neighbors and min.dist`

5 0.5

Comparative UMAP plots of the randomly selected 10000 cells from Hydra dataset under the n.neighbors 50, min.dist 0.7 and the n.neighbors 5, min.dist 0.5 optimized by scDEED:

Before optimization:

After optimization:

umap_example$`plot. # of dubious embeddings vs pair of n.neighbors and min.dist`

Plot of number of dubious embeddings vs pair of n.neighbors and min.dist for UMAP:

Example for tsne

tsne_example <- scDEED(input_counts, num_pc = 10, use_method = "tsne",visualization = TRUE)

head(tsne_example$`number of dubious cells corresponding to perplexity list`)

	perplexity	number of dubious cells
1	20	323
2	50	6
3	80	7
4	110	10
5	140	13
6	170	12

tsne_example$`best perplexity`

50

Comparative tSNE plots of the randomly selected 10000 cells from Hydra dataset under the perplexity 20 and the perplexity 50 optimized by scDEED:

Before optimization:

After optimization:

tsne_example$`plot. # of dubious embeddings vs parameters`

Plot of number of dubious embeddings vs parameters for tSNE: