Features_Preprocessing: Feature preprocessing.

In masato-ogishi/Repitope: Epitope immunogenicity prediction through in silico TCR-peptide contact potential profiling

Feature preprocessing.

Description

Features_Preprocess does preprocessing, i.e. centering and rescaling using the internally separated training subdata.
Features_CorFilter filters highly correlated features. Internally using parCor to do parallelized calculation of a correlational matrix.
Features_FeatureSelect selects important features. See also Features_Importance.
inverseColOrderDFList takes preprocessedDFList and reverses the column order of the datatable.

Usage

Features_Preprocess(featureDFList, metadataDF, seedSet = 1:5)

Features_CorFilter(
  preprocessedDFList,
  corThreshold = 0.75,
  coreN = parallel::detectCores(logical = F)
)

Features_FeatureSelect(
  preprocessedDFList,
  featureN = 100,
  coreN = parallel::detectCores(logical = F)
)

inverseColOrderDFList(preprocessedDFList)

Arguments

featureDFList	A list of feature dataframes generated by Features, or imported by readFeatureDFList.
metadataDF	A dataframe containing metadata, consisting of "Peptide", "Immunogenicity", and "Cluster" columns.
seedSet	A set of random seeds.
preprocessedDFList	A list of feature dataframes generated by Features_Preprocess or Features_CorFilter.
corThreshold	The threshold of correlation to eliminate features.
coreN	The number of cores to be used for parallelization. Set NULL to disable.
featureN	The number of features to be retained.

masato-ogishi/Repitope documentation built on Feb. 14, 2023, 5:47 a.m.