【scRNA-seq】读入hdf5格式的表达矩阵

最近尝试了好久用R读这个文件，但是hdf5r包一直出问题，于是转用python读入

import h5py
import numpy as np
import pandas as pd
import scanpy as sc
test = h5py.File("/mnt/Data16Tb/boshuang/GSE185862_expression_matrix_10x.hdf5",'r')
test.keys()#看一下里面的对象
#data下有'counts','gene','samples','shape'，分别拿出来构建Anndata
count=test['/data']['counts'][:]
cellinfo = test['/data']['samples'][:]
geneinfo = test['/data']['gene'][:]
#取出来两个都是array，要转成Dataframe
celldf = pd.DataFrame(cellinfo,index=cellinfo,columns=['cell_index'])
genedf = pd.DataFrame(geneinfo,index=geneinfo,columns=['gene_index'])
adata = sc.AnnData(count, obs=celldf, var = genedf)

至此构建AnnData成功。