The function calculates differential cell abundance the transformed proportions and performs a t-test based on a given design matrix
Usage
propellerCalc(
seu_obj1,
condition1,
condition2,
cluster_col,
meta_col,
lookup,
sample_col,
formula,
min_cells = 30
)Arguments
- seu_obj1
A seurat object
- condition1
A character representing the first condition
- condition2
A character representing the second condition
- cluster_col
A character representing the name of the column which contains cluster information
- meta_col
A character representing the name of the column which contains meta information, should be in Seurat object and in lookup table
- lookup
A dataframe that contains sample information
- sample_col
A character representing the name of the column in seu_obj1 which contains sample information
- formula
A linear model that should be used for the design matrix
- min_cells
A numeric value indicating the minimum number of cells in a cluster that should be included in the analysis
Examples
set.seed(123)
library(Seurat)
pbmc_small$condition <- factor(sample(c("diseaseA", "diseaseB"), nrow(pbmc_small), replace = TRUE))
pbmc_small$cluster <- Idents(pbmc_small)
pbmc_small$patient <- rep(paste0("P", 0:9), each = 8, length.out = ncol(pbmc_small))
lookup <- data.frame(
patient = paste0("P", 0:9),
condition = sample(c("diseaseA", "diseaseB"), 10, replace = TRUE)
)
propellerCalc(
seu_obj1 = pbmc_small,
condition1 = "diseaseA",
condition2 = "diseaseB",
cluster_col = "cluster",
meta_col = "condition",
lookup = lookup,
sample_col = "patient",
formula = "~ 0 + condition",
min_cells = 30
)
#> Performing logit transformation of proportions
#> # A tibble: 1 × 9
#> cluster PropMean.conditiondiseaseA PropMean.conditiondi…¹ PropRatio Tstatistic
#> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 0 0.667 0.357 1.87 1.28
#> # ℹ abbreviated name: ¹PropMean.conditiondiseaseB
#> # ℹ 4 more variables: P.Value <dbl>, FDR <dbl>, log2ratio <dbl>, FDR_log <dbl>