SCseq function

The SCseq Class

The SCseq Class

The SCseq class is the central object storing all information generated during cell type identification with the RaceID3 algorithm. It comprises a number of slots for a variety of objects. class

Arguments

  • object: An SCseq object.

Slots

  • expdata: The raw expression data matrix with cells as columns and genes as rows in sparse matrix format.
  • ndata: Filtered data with expression normalized to one for each cell.
  • noise: Matrix with local gene expression noise estimates (used for VarID analysis)
  • counts: Vector with total transcript counts for each cell in ndata remaining after filtering.
  • genes: Vector with gene names of all genes in ndata remaining after filtering.
  • dimRed: list object object storing information on a feature matrix obtained by dimensional reduction, batch effect correction etc. Component x stores the actual feature matrix.
  • distances: distance (or dis-similarity) matrix computed by RaceID3.
  • imputed: list with two matrices computed for imputing gene expression. The first matrix nn contains the cell indices of the knn nearest neighbours, the second matrix contains the probabilities at which each cell contributes to thye imputed gene expression value for the cell correponding to the columns.
  • tsne: data.frame with coordinates of two-dimensional tsne layout computed by RaceID3.
  • fr: data.frame with coordinates of two-dimensional Fruchterman-Rheingold graphlayout computed by RaceID3.
  • umap: data.frame with coordinates of two-dimensional umap representation computed by RaceID3.
  • cluster: list storing information on the initial clustering step of the RaceID3 algorithm
  • background: list storing the polynomial fit for the background model of gene expression variability computed by RaceID3, which is used for outlier identification.
  • out: list storing information on outlier cells used for the prediction of rare cell types by RaceID3
  • cpart: vector containing the final clustering (i.e. cell type) partition computed by RaceID3
  • fcol: vector contaning the colour scheme for the RaceID3 clusters
  • medoids: vector containing the cell ids for th cluster medoids
  • filterpar: list containing the parameters used for cell and gene filterung
  • clusterpar: list containing the parameters used for clustering
  • outlierpar: list containing the parameters used for outlier identification
  • Maintainer: Dominic Grün
  • License: GPL-3
  • Last published: 2024-11-24

Useful links