SeuObj_integration: SeuObj_integration

In yuanqingyan/singleGEO: Query and Analysis of High-throughput Single Cell Sequencing Data from GEO

SeuObj_integration

Description

Make Seurat object from the normalized data which were downloaded from GEO database. To use this function, a list containing the normalized data should be provided. This function is re-written from Seurat package.

Usage

SeuObj_integration(
  Object.list = object.list,
  Object.list2 = NULL,
  Frow.which.Norm = "lognorm",
  SampleNameAsReference = NULL,
  NumberOfSampleForReference = 4,
  NumberOfSampleForReference2 = NULL,
  Nfeatures = 2000,
  Do.scale = TRUE,
  Do.center = TRUE,
  Anchor.reduction = "rpca",
  Dims.anchor = 1:50,
  Dims.umap = 1:30,
  K.weight = 100,
  Resolution = 0.8,
  Algorithm = 1,
  Future.globals.maxSize = 12000 * 1024^2
)

Arguments

Object.list	A list containing the seurat object of each samples. This should be returned by MakeSeuObj_FromNormalizedData or MakeSeuObj_FromRawData. Required.
Object.list2	An additional list containing the seurat object of each samples for data integration. This should be returned by MakeSeuObj_FromNormalizedData or MakeSeuObj_FromRawData. Optional: NULL.
Frow.which.Norm	Which normalization method the data are obtained. Two normalization methods are allowed and supported: lognorm (For LogNormalization) and sct(For sctransform). Default: lognorm
SampleNameAsReference	A string of names to be used as the reference for data integration. The names should match the object names of the input list. Try the command of "names(object.list)" and/or "names(object.list2)" to check. Optional: NULL.
NumberOfSampleForReference	The number of objects will be used as the reference during integration. This parameter will be masked if SampleNameAsReference is not NULL. Default: 4
NumberOfSampleForReference2	The number of objects in Object.list2 will be used as the reference during integration. If this parameter is null and object.list2 is not null, it is set to be equal of SampleNameAsReference. Default: NULL
Nfeatures	The number of top variable features for the SelectIntegrationFeatures function. Default: 2000
Do.scale	Whether to scale the data or not. Binary data. This will feed to ScaleData function. Default: TRUE
Do.center	Whether to center the data or not. Binary data. This will feed to ScaleData function. Default: TRUE
Anchor.reduction	Dimensional reduction method to find anchors. Default: rpca
Dims.anchor	The number of top dimensions for the IntegrateData function. Default: 1:50
Dims.umap	The number of top dimensions of reduction to use for the functions of FindNeighbors and RunUMAP. Default: 1:30
K.weight	Number of neighbors to consider when weighting anchors in "IntegrateData" function. Default: 100
Resolution	The resolution value for FindClusters function. Default: 0.8
Algorithm	The algorithm to be used in FindClusters. Default: 1
Future.globals.maxSize	Maximum allowed total size of global variables. Default: 12000*1024^2 (12Gb)

Value

A Seurat object with data integrated.

Examples

library(singleGEO)

#
data(listDat_GSE134174)
listDat_obj<-listDat_GSE134174$data
listDat_meta<-listDat_GSE134174$meta
Int_GSE134174<-SeuObj_integration(Object.list=listDat_obj)
row.names(listDat_meta)<-listDat_meta$Cell
listDat_meta<-listDat_meta[row.names([e-mail]),]
Int_GSE134174<-Seurat::AddMetaData(object=Int_GSE134174,metadata =as.data.frame(listDat_meta))
Seurat::DimPlot(Int_GSE134174,group.by="cluster_ident",label=TRUE)

yuanqingyan/singleGEO documentation built on Jan. 11, 2025, 6:35 p.m.