ImputeMissingMetabolites: Imputes missing metabolites based on various correlation...

In jordanaron22/ImputingMetabolites: Compares R2 Value of Different Metabolite Imputation Methods

Description

Imputes missing metabolites based on different correlation matrices calculated previously. The different methods are: average correlation (ignoring heterogeneity), true correlation, MICE imputation and 1-4 factor analysis, and MICE imputation. Outputs into dataframe. Takes time (maybe a lot if large sample size) but useful for ggplot.

Usage

ImputeMissingMetabolites(
  sample_size = 1000,
  num_of_metabos = 20,
  num_to_remove = 5,
  factor_mat,
  metabolites_array_obs,
  metabolites_array_true,
  corr_avg,
  corr_array_true,
  corr_array_fa1,
  corr_array_fa2,
  corr_array_fa3,
  corr_array_fa4,
  corr_array_mice
)

Arguments

sample_size	Number of individuals generated from correlation matrices
num_of_metabos	Number of metabolites
num_to_remove	Number of metabolites removed at a time from true data in CreateCorrelation()
factor_mat	Matrix where 3 columns are three factors and rows are values of each factor
metabolites_array_obs	Metabolite data with removed metabolites generated in CreateCorrelation()
metabolites_array_true	Metabolite data generated in CreateCorrelation()
corr_avg	Avgerage correlation. Ignores heterogeneity. Output from CalcCorrAvg()
corr_array_true	True correlation matrix. An output from CreateCorrelation()
corr_array_fa1	MICE and 1 factor factor analysis. An output from FactorAnalysis()
corr_array_fa2	MICE and 2 factor factor analysis. An output from FactorAnalysis()
corr_array_fa3	MICE and 3 factor factor analysis. An output from FactorAnalysis()
corr_array_fa4	MICE and 4 factor factor analysis. An output from FactorAnalysis()
corr_array_mice	Imputed with MICE. First output from ReorganizeAndMICE()

Value

data frame of correlation values by metabolite, factor, and imputation type

jordanaron22/ImputingMetabolites documentation built on Dec. 21, 2021, 2:18 a.m.