cluster_mat: matplot per cluster

In shkonishi/cornet: Cunstruction of correlation network form transcriptome data

Description

split count data of transcriptome with result of gene clustering, such as "cutree" or "cutreeDynamic".

Usage

cluster_mat(dat, res_dycut, fcdat, fctrl_col, facet_ncol, facet_nrow)

Arguments

dat	a data frame of count data for transcriptome, which columns are genes and row are samples.
res_dycut	result of "dynamicTreeCut::cutreeDynamic", or "cutree"
fcdat	data.frame of one or two factor for samples, which first column is a factor for Y-axis and a second column is a factor for X-axis. If samples factor is not unique and replicate, median of replicate was expression value.
fctrl_col	colour code for colour factor. The default value is NULL, then default ggplot colours.
facet_ncol	ncol argument of facet_wrap. The default value is 3.
facet_nrow	nrow argument of facet_wrap. The default value is 4

Examples

#
# # argument1: data containing replicate or not.
# nfpkm <- rskodat::nfpkm
# nfpkm_norep <- nfpkm %>% group_by_at(1:4) %>% summarise_at(vars(-1:-4), funs(median))
#
# # argument2: result of dycutdf
# res_dyct1 <- cornet::dycutdf(nfpkm, distm = "spearman", column = -1:-4)
# res_dyct2 <- cornet::dycutdf(nfpkm_norep, distm = "spearman", column = -1:-4)
#
# cl1 <- res_dyct1$dynamic_cut
# cl2 <- res_dyct2$dynamic_cut
#
# # argument3: factorial data.frame one(x) or two(y,x)
# fcdat1 <- data.frame(runs=nfpkm$runs, days=nfpkm$days)
# fcdat2 <- data.frame(runs=nfpkm_norep$runs, days=nfpkm_norep$days)
#
# # execution
# res1 <- cornet::cluster_mat(dat = nfpkm[-1:-4], res_dycut = cl1, fcdat = fcdat1)
# res2 <- cornet::cluster_mat(dat = nfpkm_norep[-1:-4], res_dycut = cl2, fcdat = fcdat2,
#                            fctrl_col = c(1,2,4))
#
# # drawing selected panels
# do.call(gridExtra::grid.arrange, c(res2[c(2,5)], list(ncol=2)))

shkonishi/cornet documentation built on May 30, 2019, 7:09 p.m.