###
# Compute performance measures (correlation, mean squared error).
###
print(snakemake@input[[1]])
mse_with_deconf = c()
cor_with_deconf = c()
mse_no_deconf = c()
cor_no_deconf = c()
for(file in snakemake@input){
out = mget(load(file, envir=(NE. <- new.env())), envir=NE.)
x1 = out$normal_fit_with_deconf$dce_pvalue
x2 = out$extended_fit_with_deconf$dce_pvalue
shared_genes = intersect(colnames(x1), colnames(x2))
x1 = x1[shared_genes, shared_genes]
x2 = x2[shared_genes, shared_genes]
x1 = -log(c(x1)[is.na(c(x1))==FALSE] + 1e-10)
x2 = -log(c(x2)[is.na(c(x2))==FALSE] + 1e-10)
mse_with_deconf = c(mse_with_deconf, mean((x1-x2)^2))
cor_with_deconf = c(cor_with_deconf, cor(x1, x2))
y1 = out$normal_fit_no_deconf$dce_pvalue
y2 = out$extended_fit_no_deconf$dce_pvalue
shared_genes = intersect(colnames(y1), colnames(y2))
y1 = y1[shared_genes, shared_genes]
y2 = y2[shared_genes, shared_genes]
y1 = -log(c(y1)[is.na(c(y1))==FALSE] + 1e-10)
y2 = -log(c(y2)[is.na(c(y2))==FALSE] + 1e-10)
mse_no_deconf = c(mse_no_deconf, mean((y1-y2)^2))
cor_no_deconf = c(cor_no_deconf, cor(y1, y2))
}
save(mse_with_deconf,
cor_with_deconf,
mse_no_deconf,
cor_no_deconf,
file=snakemake@output[[1]])
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.