Find the hit using the roast method. Roast is a competitive gene set test which uses rotation instead of permutation. Here is applied in a contest of a genetic screening so it perform a barcode competitive test testing for barcode which are differentially expressed within a gene. More information can be found in Roast
find_roast_hit(
screenR_Object,
matrix_model,
contrast,
nrot = 9999,
number_barcode = 3,
direction = "Down",
p_val = 0.05
)The ScreenR object obtained using the
create_screenr_object
The matrix that will be used to perform the
linear model analysis. Created using
model.matrix
A vector or a single value indicating the index or the name of the column the model_matrix to which perform the analysis
Number of rotation to perform the test. Higher number of rotation leads to more statistically significant result.
Number of barcode that as to be differentially expressed (DE)in order to consider the gene associated DE. Example a gene is associated with 10 shRNA we consider a gene DE if it has at least number_barcode = 5 shRNA DE.
Direction of variation
The value that as to be used as p-value cut off
The hits found by ROAST method
set.seed(42)
object <- get0("object", envir = asNamespace("ScreenR"))
matrix_model <- model.matrix(~ slot(object, "groups"))
colnames(matrix_model) <- c("Control", "T1_T2", "Treated")
result <- find_roast_hit(object,
matrix_model = matrix_model,
contrast = "Treated", nrot = 100
)
#> Warning: 3rows with all zero counts
head(result)
#> # A tibble: 6 × 9
#> Gene NGenes PropDown PropUp Direction PValue FDR PValue.Mixed FDR.Mixed
#> <chr> <int> <dbl> <dbl> <fct> <dbl> <dbl> <dbl> <dbl>
#> 1 Gene_173 10 0.7 0 Down 0.00990 0.264 0.00990 0.610
#> 2 Gene_15 10 0.4 0 Down 0.00990 0.264 0.00990 0.610
#> 3 Gene_121 10 0.4 0 Down 0.00990 0.264 0.00990 0.610
#> 4 Gene_293 10 0.4 0.1 Down 0.00990 0.264 0.0198 0.610
#> 5 Gene_394 10 0.3 0 Down 0.00990 0.264 0.366 0.846
#> 6 Gene_457 10 0.2 0 Down 0.00990 0.264 0.455 0.846