SDImpute: SDImpute: A statistical block imputation method based on...

In Jinsl-lab/SDImpute: SDImpute: A statistical block imputation method based on cell-level and gene-level information for dropouts in single-cell RNA-seq data

Description

SDImpute: A statistical block imputation method based on cell-level and gene-level information for dropouts in single-cell RNA-seq data

Usage

SDImpute(
  data,
  do.nor = TRUE,
  auto.k = TRUE,
  criterion = "asw",
  krange = c(5:15),
  k = 5,
  M = 15,
  T = 0.5
)

Arguments

data	A gene expression matrix,the rows correspond to genes and the columns correspond to cells.
do.nor	Logical. If TRUE, the data is Normalized.
auto.k	Logical. If TRUE, k is estimated by either the Calinski Harabasz index or average silhouette width ; If FALSE, the parameter k need to be set manually.
criterion	One of "asw" or "ch". Determines whether average silhouette width or Calinski-Harabasz is applied.
krange	Integer vector. Numbers of clusters which are to be compared by the average silhouette width criterion. Note: average silhouette width and Calinski-Harabasz can't estimate number of clusters nc=1.
k	Integer. The number of cell clusters. This parameter can be determined based on prior knowledge or clustering result of raw data.
M	Integer. The number of nearest neighbors.When the number of nearest neighbors for each cell is small, the parameter M should not be too large to guarantee that it makes sense. In general, this parameter is set to an integer between 10 and 30.
T	Numeric between 0 and 1. The dropout probability candidate threshold which controls the degree of imputation to the gene expression matrix. The recommended value of parameter T is 0.5.

Value

An imputation matrix

Examples

library("SDImpute")
data(data)
imputed_data<-SDImpute(data,do.nor=TRUE,auto.k=FALSE,k=5,M=15,T=0.5)

Jinsl-lab/SDImpute documentation built on June 9, 2021, 9:12 a.m.