extract_centroids: Extract tandem repeat unit squiggle centroids.

In arnederoeck/NanoSatelliteR: QC, plotting, and clustering of tandem repeat segmentation created by NanoSatellite.

Description

extract_centroids() obtains the squiggle current levels which best represent their respective clusters.

Usage

extract_centroids(tsclust_obj)

Arguments

tsclust_obj

A tsclust object generated by dtwclust::tsclust()

Details

dtwclust::tsclust() can be used to cluster tandem repeat unit squiggles and centroids are calculated for each cluster. This centroid aims to provide the the most representative squiggle current level for each cluster. extract_centroids extracts the centroid information from a tsclust object and returns a data.frame with the results

Examples

df <- load_summary("/storage/NanoSatellite_chunks/")
qc <- summary_qc(df)
df2 <- qual_reads(df,qc$center_cutoff)
squiggles <- load_squiggles("/storage/NanoSatellite_chunks/",df2)

library(doParallel)
library(dtwclust)
k_clusters=2
registerDoParallel(cores=8)

positive_clustering=tsclust(squiggles$positive,type="h",k=k_clusters,trace=TRUE,distance = "dtw_basic", control=hierarchical_control(method="ward.D",symmetric = T))

cent=extract_centroids(positive_clustering)
ggplot(cent,aes(x=pos,y=signal,colour=factor(cluster)))+geom_point()+geom_line()+theme_minimal()+facet_grid(. ~ cluster)+guides(colour=guide_legend(title="cluster"))

arnederoeck/NanoSatelliteR documentation built on May 6, 2019, 1:35 a.m.