plot_pca: Run and visualise PCA

In Meiomap/SureTypeSCR: Library for (single-cell) SNP array processing

Description

Function creates feature matrix from non-zero SNPs and feature types indicated in features. The feature matrix used for calculating PCA has n rows and k columns, while n corresponds to number of individuals in the data frame and k is defined as length of features x number of non-zero SNPs across all samples.

Usage

plot_pca(.data,by_chrom=TRUE,features=c('x','y'),importances=TRUE,metadata=NULL,labels=TRUE)

Arguments

.data	dataframe from scbasic or compatible
by_chrom	boolean flag indicating whether to run PCA by chromosome (TRUE or FALSE)
features	vector of feature types to use for PCA (columns from .data , i.e. c('x','y'))
importances	boolean flag indicating whether to render dimension importances (TRUE or FALSE)
metadata	data frame with metadata - currently grouping by family is supported, given that metadata contains familyid and individual columns and metadata will be merged to internal data frame by matching individual
labels	boolean flag indicating whether to render sample names using package ggrepel to prevent overplotting (TRUE or FALSE)

Value

ggplot object with PCA plot

Examples

setwd(system.file(package='SureTypeSCR'))

samplesheet=system.file('files/GSE19247_example.csv',package='SureTypeSCR')
manifest=system.file('files/HumanCytoSNP-12v2_H.bpm',package='SureTypeSCR')
cluster=system.file('files/HumanCytoSNP-12v2_H.egt',package='SureTypeSCR')

#Load data
df=scbasic(manifest,cluster,samplesheet)

#create ggplot object with PCA
#df %>% plot_pca(by_chrom=TRUE,features=c('x','y'),labels=TRUE)

Meiomap/SureTypeSCR documentation built on Dec. 17, 2021, 3:22 a.m.