ssvSignalClustering {seqsetvis} | R Documentation |
ssvSignalHeatmap
but can also be run before calling
ssvSignalHeatmap for greater control and access to clustering results
directly.Clustering is via k-means by default. The number of clusters is determined
by nclust. Optionally, k-means can be initialized with a data.frame provided
to k_centroids. As an alternative to k-means, a membership table from
ssvMakeMembTable
can be provided to determine logical clusters.
ssvSignalClustering( bw_data, nclust = NULL, k_centroids = NULL, memb_table = NULL, row_ = "id", column_ = "x", fill_ = "y", facet_ = "sample", cluster_ = "cluster_id", max_rows = 500, max_cols = 100, clustering_col_min = -Inf, clustering_col_max = Inf, within_order_strategy = c("hclust", "sort")[2], dcast_fill = NA )
bw_data |
a GRanges or data.table of bigwig signal. As returned from
|
nclust |
Number of clusters. Defaults to 6 if nclust, k_centroids, and memb_table are not set. |
k_centroids |
data.frame of centroids for k-means clusters. Incompatible with nclust or memb_table. |
memb_table |
Membership table as from |
row_ |
variable name mapped to row, likely id or gene name for ngs data. Default is "id" and works with ssvFetch* output. |
column_ |
varaible mapped to column, likely bp position for ngs data. Default is "x" and works with ssvFetch* output. |
fill_ |
numeric variable to map to fill. Default is "y" and works with ssvFetch* output. |
facet_ |
variable name to facet horizontally by. Default is "sample" and works with ssvFetch* output. Set to "" if data is not facetted. |
cluster_ |
variable name to use for cluster info. Default is "cluster_id". |
max_rows |
for speed rows are sampled to 500 by default, use Inf to plot full data |
max_cols |
for speed columns are sampled to 100 by default, use Inf to plot full data |
clustering_col_min |
numeric minimum for col range considered when clustering, default in -Inf |
clustering_col_max |
numeric maximum for col range considered when clustering, default in Inf |
within_order_strategy |
one of "hclust" or "sort". if hclust, hierarchical clustering will be used. if sort, a simple decreasing sort of rosSums. |
dcast_fill |
value to supply to dcast fill argument. default is NA. |
Within each cluster, items will either be sorted by decreasing average signal or hierachically clustered; this is controlled via within_order_strategy.
data.table of signal profiles, ready for ssvSignalHeatmap
clust_dt = ssvSignalClustering(CTCF_in_10a_profiles_gr) ssvSignalHeatmap(clust_dt) clust_dt2 = ssvSignalClustering(CTCF_in_10a_profiles_gr, nclust = 2) ssvSignalHeatmap(clust_dt2) #clustering can be targetted to a specific part of the region clust_dt3 = ssvSignalClustering(CTCF_in_10a_profiles_gr, nclust = 2, clustering_col_min = -250, clustering_col_max = -150) ssvSignalHeatmap(clust_dt3) clust_dt4 = ssvSignalClustering(CTCF_in_10a_profiles_gr, nclust = 2, clustering_col_min = 150, clustering_col_max = 250) ssvSignalHeatmap(clust_dt4)