LEAP (version 0.2)

MAC_perm: Function to perform a permutation analysis to determine a cutoff for significant MAC values.

Description

See vignette for more details.

Usage

MAC_perm(data, MACs_observ, num_perms = 100, max_lag_prop = 1/3,  FDR_cutoffs = 101, perm_file_name = F)

Arguments

data

A data matrix for which the rows are genes and the columns are experiments, sorted by their pseudotime.

MACs_observ

The resulting matrix of MACs from running MAC_counter on the dataset

num_perms

The number of permutations to be performed. Default is 100.

max_lag_prop

The largest proportion of your experiments that you want the size of the lag to be. Recommended not to go beyond 1/3. Default value is 1/3.

FDR_cutoffs

The number of cutoffs between 0 and 1 to use for FDR analysis. Default value is 101, resulting in 0,0.01,0.02,...,0.98,0.99,1.

perm_file_name

The name to be used for resulting .csv file. I.e., using perm_file_name="mine" would create the file "perm_mine.csv"

Value

Returns a dataset with four columns: cors are the correlation cutoffs, MACs_observed are the number of observed correlations at that cutoff, MACs_ave_perm are the average number observed in the permuted datasets at that cutoff, and fdr is the false discovery rate (FDR) observed at that cutoff.

Details

See vignette for more details.

Examples

x <- matrix(rnorm(6),2,3)
cor <- cor(x)
y <- MAC_perm(x, cor)

## Not run: MAC_perm(data=example_data, MACs_observ=MAC_example)