scPred: Definition of 'scPred' class

In powellgenomicslab/scPred: scPred - accurate supervised method for cell-type classification from single-cell RNA-seq data

Description

An S4 class to containing features, dimensionality reduction information, and trained models

Slots

pvar

Column name from metadata to use as the variable to predict using

metadata

A data frame to store metadata including prediction variable (pvar)

features

A data frame with the following information:

• feature: Eigenvector (e.g. principal component)

• pValue: Significance value from a wilcoxon test

• pValueAdj: Adjusted p-value for multiple testing

loadings

Gene loadings

scaling

Means and standard deviation to center and standardize data

reduction

Dimensionality reduction name

reduction_key

Dimensionality reduction name key

train

A list with all trained models using the caret package. Each model correspond to a cell type

mist

A list to store extra information and for developing testing

powellgenomicslab/scPred documentation built on July 16, 2021, 12:14 a.m.