通用代码：python中的h5、h5ad文件批量转为seuart对象 通用代码：python中的h5、h5ad文件批量转为seu

import scanpy as scimport os​# Define the directory containing the h5 filesdirectory = '/home/data/t040413/20240125_kidney'​# List all h5 files in the directoryh5_files = [f for f in os.listdir(directory) if f.endswith('.h5')]print(h5_files)​# Initialize an empty list to store the AnnData objectsadata_list = []​# Read each h5 file and append the AnnData object to the listfor file in h5_files:    file_path = os.path.join(directory, file)    adata = sc.read_10x_h5(file_path)    adata.var_names_make_unique()        # Extract the sample name from the file name    sample_name = file.replace("_filtered_feature_bc_matrix.h5", "")    print(sample_name)    adata.obs['sample'] = sample_name  # Assign sample name to each cell        # Basic filtering    sc.pp.filter_cells(adata, min_genes=200)    sc.pp.filter_genes(adata, min_cells=3)​    adata_list.append(adata)​# Concatenate all AnnData objects into oneall_data = adata_list[0].concatenate(adata_list[1:], batch_key='Sample')#batch_key 只是添加一个标签而已，真正的分组信息在最开始就弄好了​# Print the concatenated dataprint(all_data)​

import scipy.sparse as sparseimport scipy.io as sioimport scipy.stats as statsimport numpy as npcellinfo=all_data.obsgeneinfo=all_data.varmtx=all_data.X.T​​​cellinfo.to_csv("cellinfo.csv")geneinfo.to_csv("geneinfo.csv")sio.mmwrite("sparse_matrix.mtx",mtx)!pwd

cellinfo=read.csv("./cellinfo.csv",row.names = "X")head(cellinfo)​geneinfo=read.csv("./geneinfo.csv",row.names = "X");head(geneinfo)geneinfo=geneinfo[,c(2,3)]head(geneinfo)​# counts=ReadMtx(mtx = "./sparse_matrix.mtx",#                cells = "./cellinfo.csv",#                features = "./geneinfo.csv")​counts=Matrix::readMM(file = "./sparse_matrix.mtx")head(counts)[,1:9]dim(counts)rownames(counts)=rownames(geneinfo)colnames(counts)=rownames(cellinfo)library(Seurat)kidney=CreateSeuratObject(counts = counts,project = "kidney",meta.data = cellinfo)dim(kidney) #​