scVelo

一.導入需要的文件

需要的文件：由R裏面對Seurat對象的數據導出的
1.velocyto pipeline 跑出來的loom文件
2.細胞名字文件
3.細胞屬於的類羣

#RNA剪切信息的loom文件
sample_one = anndata.read_loom("NC.loom")
#包含細胞名的data.frame
sample_obs = pd.read_csv("NC_cellID_obs.csv")
#包含細胞類羣註釋信息的數據框
cell_clusters = pd.read_csv("/NC_clusters.csv")

二.更改文件細胞名字，使其一致：

1.對於loom文件

sample_one.obs

#去除NC和x只留barcode名字
sample_one.obs=sample_one.obs.rename(index = lambda x: x.replace('NC:', ''))
sample_one.obs=sample_one.obs.rename(index = lambda x: x.replace('x', ''))
sample_one.obs.head()

2.對於細胞名

sample_obs.head()

#去除barcode的_1
sample_obs.x=sample_obs.x.replace({"_1":""},regex=True)
sample_obs.head()

3.umap座標文件的細胞名字更改

umap.head()

umap[["Unnamed: 0"]]=umap[["Unnamed: 0"]].replace({"_1":""},regex=True)
umap = umap.rename(columns = {"Unnamed: 0":'Cell ID'})
umap.head()

4.對於細胞類羣文件

cell_clusters.head()

cell_clusters[["Unnamed: 0"]]=cell_clusters[["Unnamed: 0"]].replace({"_1":""},regex=True)
cell_clusters.head()

三.對細胞進行過濾並排序

#對細胞文件和RNA剪切速率文件取交集
sample_one = sample_one[np.isin(sample_one.obs.index,sample_obs["x"])]
sample_one.obs.head()
#對UMAP座標文件取交集
sample_one_index = pd.DataFrame(sample_one.obs.index)
sample_one_index = sample_one_index.rename(columns = {0:'Cell ID'})
umap_ordered = sample_one_index.merge(umap,on="Cell ID")
umap_ordered.head()
#對 細胞類羣文件細胞取交集
cell_clusters[["Unnamed: 0"]]=cell_clusters[["Unnamed: 0"]].replace({"_1":""},regex=True)
cell_clusters = cell_clusters.rename(columns = {"Unnamed: 0":'Cell ID'})
#order
cell_clusters = sample_one_index.merge(cell_clusters,on="Cell ID")
cell_clusters.head()

四.將umap座標與cluster信息加入sample_one

#將umap信息加入sample_one
umap_ordered = umap_ordered.iloc[:,1:]
sample_one.obsm['X_umap'] = umap_ordered.values
#將cell_clusters加入sample_one
cell_clusters_ordered=cell_clusters.iloc[:,2]
sample_one.obs['cell_clusters']=cell_clusters_ordered.values

五.運行RNA Velocity

#Running RNA Velocity
scv.pp.filter_and_normalize(sample_one,min_shared_counts=30, n_top_genes=2000)
scv.pp.moments(sample_one, n_pcs=30, n_neighbors=30)
scv.tl.velocity(sample_one, mode = "stochastic")
scv.tl.velocity_graph(sample_one)
scv.pl.velocity_embedding(sample_one, basis='X_umap',arrow_size=5)
ident_colours = ["#F8766D","#7CAE00","#00BFC4","#C77CFF"]
scv.pl.velocity_embedding_stream(sample_one, basis='X_umap',color = "cell_clusters",palette = ident_colours)