Screen feature-phenotype correlation

Use metadata-provided variables (any phenotypes or environmental variables) to screen feature-phenotype correlation linear model, linear mixed model (using groups as random effect), or correlation (Pearson, Spearman, Kendall) are supported

Usage

ScreenFeaturePhenotypeCorrelation(
  mmo,
  phenotype,
  groups,
  model = "lm",
  normalization = "None"
)

Arguments

mmo

The mmo object with feature data and metadata

phenotype

The name of the phenotype in the metadata

groups

A vector of group names from the metadata containing phenotype data

model

The type of regression model to use. Options are 'lmm' for linear mixed model, 'lm' for simple linear regression, or 'pearson', 'spearman', 'kendall' for correlation (default: 'lm')

normalization

The normalization method to use for feature data. Options are 'None', 'Log', 'Meancentered', or 'Z' (default: 'Z')

Value

A list of the plot and the raw data