想要將scanpy的h5ad文件轉為rds文件，有一個很方便的包sceasy可以使用，安裝簡單，代碼簡易。

安裝方式

# bioconda package:conda install -c bioconda r-sceasy# R package:devtools::install_github("cellgeni/sceasy")#額外需要安裝的依賴包
if (!requireNamespace("BiocManager", quietly = TRUE))install.packages("BiocManager")BiocManager::install(c("LoomExperiment", "SingleCellExperiment"))install.packages('reticulate')

R語言腳本

Seurat to AnnData

sceasy::convertFormat(seurat_object, from="seurat", to="anndata",outFile='filename.h5ad')

?AnnData to Seurat

sceasy::convertFormat(h5ad_file, from="anndata", to="seurat",outFile='filename.rds')

Seurat to SingleCellExperiment

sceasy::convertFormat(seurat_object, from="seurat", to="sce",outFile='filename.rds')

SingleCellExperiment to AnnData?

sceasy::convertFormat(sce_object, from="sce", to="anndata",outFile='filename.h5ad')

SingleCellExperiment to Loom?

sceasy::convertFormat(sce_object, from="sce", to="loom",outFile='filename.loom')

?Loom to AnnData

sceasy::convertFormat('filename.loom', from="loom", to="anndata",outFile='filename.h5ad')

?Loom to SingleCellExperiment

sceasy::convertFormat('filename.loom', from="loom", to="sce",outFile='filename.rds')

批量處理的代碼?

library(Seurat)
library(stringr)
library(sceasy)
library(reticulate)
loompy <- reticulate::import('loompy')use_condaenv('EnvironmentName') #使用的python路徑或者conda中python的路徑#h5ad轉rds文件，文件夾下所有文件批量轉換代碼outdir = 'XXX/Glycine.max/'
setwd(outdir)
file_names <- list.files(path = outdir,  pattern = "\\.h5ad$", full.names = FALSE)
for (i in 1:length(file_names)) 
{sample1 <- file_names[i]sample2 <- sub("\\..*", "", file_names[i])print(sample1)h5ad_file=str_c(sample1)  #輸入的h5文件路徑out_file=str_c(outdir2,sample2,".rds")sceasy::convertFormat(h5ad_file, from="anndata", to="seurat", outFile=out_file)
}

記住需要添加use_condaenv的python路徑，不然會使用默認的python，容易報錯。報錯如下圖所示。use_condaenv('XXX/bin/python')

Error: PyType_GetFlags - /usr/lib64/python2.7/config/libpython2.7.so: undefined symbol: PyType_GetFlags
In addition: Warning message:
In py_initialize(config$python, config$libpython, config$pythonhome,  :Python 2 reached EOL on January 1, 2020. Python 2 compatability will be removed in an upcoming reticulaterelease.

?