get_exp_cor_edges {PFP} | R Documentation |
compute the correlation coefficient of gene expression data, return the most related genes
get_exp_cor_edges( gene_list, data_std, method = "spearman", num = 5, cor_threshold = NULL )
gene_list, |
a vector of characters |
data_std, |
a matrix of data, such as gene expression data, whose rownames are gene names or ids and colnames are sample names |
method, |
a chareater, which method to compare the correlation of gene expression data it could be "pearson", "kendall", "spearman", "spearman" is default |
num, |
an integer, the top number of co-expressed genes to choose, 5 is default |
cor_threshold, |
a numeric, the threshold of the correlation coefficient to choose, default is NULL |
This function computes the correlation coefficient of gene
expression data between gene_list
and data_std
, it will return a data.frame which can be translated
a graph or network.
In the data.frame, source
refers to the genes in gene_list
,
target
refers to the top coexpressed genes,
weight
refers to the correlated coefficient of genes in source
and target
, pathway
is "uncertain" and edge_type
is
"coexp".Note, when choosing the top co-expressed genes, we will use the
num
param if the cor_threshold
param is NULL. If not,
we will choose the cor_threshold
param.
the coexp of edges.
data(data_std) data(PFP_test1) rank1 <- rank_PFP(object = PFP_test1,total_rank = TRUE) pathway_select <- refnet_info(rank1)[1,"id"] gene_test <- pathways_score(rank1)$genes_score[[pathway_select]]$ENTREZID edges_coexp <- get_exp_cor_edges(gene_test,data_std)