discoQC: Quality Control for DiscoRhythm

In matthewcarlucci/DiscoRhythm: Interactive Workflow for Discovering Rhythmicity in Biological Data

Quality Control for DiscoRhythm

Description

Functions for executing outlier detection and row filtering procedures prior to rhythmicity analysis.

Usage

discoPCAoutliers(se, threshold = 3, scale = TRUE, pcToCut = seq_len(4))

discoInterCorOutliers(
  se,
  cor_method = c("pearson", "kendall", "spearman"),
  threshold = 3,
  thresh_type = c("sd", "value")
)

discoRepAnalysis(
  se,
  aov_method = c("Equal Variance", "Welch", "None"),
  aov_pcut = 0.05,
  aov_Fcut = 0,
  avg_method = c("Median", "Mean", "Random", "None")
)

Arguments

se	SummarizedExperiment, the main data object used by DiscoRhythm expected to contain se$ID, se$ReplicateID, se$Time sample metadata and non-null rownames. See the vignette for more details.
threshold	numeric, a threshold determining which samples are outliers (for discoInterCorOutliers, in units of thresh_type, for discoPCAoutliers in units of standard deviations).
scale	logical, whether or not to scale the data prior to PCA, see prcomp for more details.
pcToCut	numeric, which PCs to use for outlier detection. It is recomended to select the first X PCs based on which PCs explain a significant amount of variance in the data.
cor_method	character, method of pairwise correlation (see cor's "method" argument for all options).
thresh_type	character indicating threshold type (either standard deviations below the mean, or an absolution correlation value). One of: "sd" or "value".
aov_method	character, method to use for ANOVA. One of: "Equal Variance", "Welch", or "None".
aov_pcut	numeric, p-value cutoff used to select rows with statistically significant signal-to-noise.
aov_Fcut	numeric, F-statistic cutoff used to select rows with high signal-to-noise based on magnitude.
avg_method	character, method for averaging technical replicates. One of: "Median","Mean","Random", or "None".

Value

list containing PCA results and the detected outliers

A list of 3 objects: 1) outliers - named logical indicating if the sample is an outlier 2) meanCor - mean of all pairwise correlations for a given sample 3) corMat - Matrix of all pairwise correlation values

Examples

se <- discoGetSimu(TRUE)
PCAres <- discoPCAoutliers(se)

CorRes <- discoInterCorOutliers(se)

ANOVAres <- discoRepAnalysis(se)

matthewcarlucci/DiscoRhythm documentation built on March 24, 2022, 10:22 a.m.