PredictAssay: Predict value from nearest neighbors

In Seurat: Tools for Single Cell Genomics

Predict value from nearest neighbors

Description

This function will predict expression or cell embeddings from its k nearest neighbors index. For each cell, it will average its k neighbors value to get its new imputed value. It can average expression value in assays and cell embeddings from dimensional reductions.

Usage

PredictAssay(
  object,
  nn.idx,
  assay,
  reduction = NULL,
  dims = NULL,
  return.assay = TRUE,
  slot = "scale.data",
  features = NULL,
  mean.function = rowMeans,
  seed = 4273,
  verbose = TRUE
)

Arguments

object	The object used to calculate knn
nn.idx	k near neighbour indices. A cells x k matrix.
assay	Assay used for prediction
reduction	Cell embedding of the reduction used for prediction
dims	Number of dimensions of cell embedding
return.assay	Return an assay or a predicted matrix
slot	slot used for prediction
features	features used for prediction
mean.function	the function used to calculate row mean
seed	Sets the random seed to check if the nearest neighbor is query cell
verbose	Print progress

Value

return an assay containing predicted expression value in the data slot

Seurat documentation built on May 29, 2024, 4:20 a.m.