{
"cells": [
{
"cell_type": "markdown",
"id": "e02e2afa",
"metadata": {},
"source": [
"# Cell transfer across ST slides of multiple time points"
]
},
{
"cell_type": "markdown",
"id": "4b54493b",
"metadata": {},
"source": [
"Multiple ST tissue sections obtained at different timepoints offer a valuable opportunity to gain a deeper understanding of cell transitions during organ development, such as mouse midbrain development and zebrafish embryogenesis. spaTrack can provide a novel strategy to infer cell tracing across multiple tissue sections based on an unbalanced and structured optimal transport algorithm by direct mapping cells among samples without integrating data. This approach is designed for cases where multiple sections containing the same cell types.In this tutorial, We use mouse midbrain ST data to demonstrate cell transfer across ST slides of multiple time points."
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "2c754320",
"metadata": {},
"outputs": [],
"source": [
"import warnings\n",
"\n",
"warnings.filterwarnings(\"ignore\")"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "26e24976-e691-4d8a-8b45-477d70825aad",
"metadata": {},
"outputs": [],
"source": [
"import spaTrack as spt\n",
"import scanpy as sc\n",
"import pandas as pd\n",
"import matplotlib.pyplot as plt\n",
"\n",
"sc.settings.verbosity = 0\n",
"plt.rcParams['figure.dpi'] = 200 #分辨率"
]
},
{
"cell_type": "markdown",
"id": "402f520d",
"metadata": {},
"source": [
"## 1.Prepare ST data"
]
},
{
"cell_type": "markdown",
"id": "33cfa5d5",
"metadata": {},
"source": [
"Spatial transcriptmic data of mouse midbrain at E12.5,E14.5 and E16.5 time points were downloaded from published paper(Chen et al.2022, Cell). The example of expression tsv file and additional information file in this part can be download from [Baidu Netdisk](https://pan.baidu.com/s/1YzBecoNb2Yu1cVF77Ly3ig?pwd=u1ty) and [Google Drive](https://drive.google.com/drive/folders/1fA-tLB1bHcAkk-lxwC3laAGpgIJfnnuD?usp=drive_link) .We focused on progenitors cells including RGC(radial glia cells),NeuB(neuroblasts) and GlioB(glioblasts) to describe transfer relations among these progenitors cells."
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "11848dd4",
"metadata": {},
"outputs": [],
"source": [
"adata_all=sc.read('../../../../data/mouse.midbrain.three.time/mouse midbrain.three.time.point.ST.exp.tsv')\n",
"df_annot=pd.read_table('../../../../data/mouse.midbrain.three.time/meta.tsv',sep='\\t',index_col=[0])\n",
"adata_all.obs=df_annot\n",
"adata_all.obs.index=list(df_annot.index)\n",
"adata_all.obsm['spatial']=adata_all.obs[['x','y']].values\n",
"adata_all.obs['CellID']=adata_all.obs.index"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "1851b0ad",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" annotation \n",
" Time point \n",
" x \n",
" y \n",
" batch \n",
" CellID \n",
" \n",
" \n",
" \n",
" \n",
" CELL.100034_1 \n",
" RGC \n",
" E12.5 \n",
" 2829.0 \n",
" -10000.0 \n",
" 0 \n",
" CELL.100034_1 \n",
" \n",
" \n",
" CELL.100795_1 \n",
" RGC \n",
" E12.5 \n",
" 2893.0 \n",
" -9946.5 \n",
" 0 \n",
" CELL.100795_1 \n",
" \n",
" \n",
" CELL.100841_1 \n",
" RGC \n",
" E12.5 \n",
" 2672.0 \n",
" -9944.0 \n",
" 0 \n",
" CELL.100841_1 \n",
" \n",
" \n",
" CELL.100867_1 \n",
" RGC \n",
" E12.5 \n",
" 2728.0 \n",
" -9943.0 \n",
" 0 \n",
" CELL.100867_1 \n",
" \n",
" \n",
" CELL.100880_1 \n",
" RGC \n",
" E12.5 \n",
" 2927.0 \n",
" -9939.0 \n",
" 0 \n",
" CELL.100880_1 \n",
" \n",
" \n",
"
\n",
"
"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"fig, axs = plt.subplots(ncols=1, nrows=1, figsize=(12, 6))\n",
"sc.pl.embedding(adata_all, basis='spatial', color='annotation', size=50, legend_loc='right margin', legend_fontsize=20,ax=axs, legend_fontoutline=3, show=False, s=80)\n",
"axs.set_title('spatial',fontsize=18)"
]
},
{
"cell_type": "markdown",
"id": "424da25f",
"metadata": {},
"source": [
"## 2.Calculate transition probability across sections"
]
},
{
"cell_type": "markdown",
"id": "a174dfa9",
"metadata": {},
"source": [
" spaTrack infers cell tracing across multiple ST sections based on an unbalanced and structured optimal transport algorithm. We can calculate the transfer matrix between two ST sections using the spt.transfer_matrix function. Additionally, the transition probability between two sections can be calculated using spt.map_data. To obtain complete results across multiple sections, spt.generate_animate_input was used to integrate the transition probability results.\n"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "81b7f963-4e0f-49d5-a40e-f228320bb586",
"metadata": {},
"outputs": [],
"source": [
"pi12 = spt.transfer_matrix(adata1_sub,adata2_sub,epsilon=0.01,alpha=0.5)\n",
"pi23 = spt.transfer_matrix(adata2_sub,adata3_sub,epsilon=0.01,alpha=0.5)"
]
},
{
"cell_type": "markdown",
"id": "9d73d142",
"metadata": {},
"source": [
"We also can obtain the result containing cell transition probability among three ST sections."
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "ea1cc498-9908-4b0f-8409-a6219addf83c",
"metadata": {},
"outputs": [],
"source": [
"pi_matrix_coord = spt.generate_animate_input(pi_list=[pi12,pi23], adata_list=[adata1_sub,adata2_sub,adata3_sub],time ='Time point',annotation='annotation')\n",
"# pi_matrix_coord=pi_matrix_coord.loc[(pi_matrix_coord['pi_value1']> value1) & (pi_matrix_coord['pi_value1']> value2) ]\n",
"pi_matrix_coord.to_csv(\"./cell.transfer.csv\")"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "327bb1e2",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" slice1 \n",
" slice2 \n",
" slice3 \n",
" slice1_annotation \n",
" slice2_annotation \n",
" slice3_annotation \n",
" slice1_x \n",
" slice2_x \n",
" slice3_x \n",
" slice1_y \n",
" slice2_y \n",
" slice3_y \n",
" k_line_12 \n",
" b_line_12 \n",
" k_line_23 \n",
" b_line_23 \n",
" pi_value1 \n",
" pi_value2 \n",
" \n",
" \n",
" \n",
" \n",
" 0 \n",
" CELL.100034_1 \n",
" CELL.174282_3 \n",
" CELL.6090_2 \n",
" RGC \n",
" GlioB \n",
" RGC \n",
" 2829.0 \n",
" 7709.0 \n",
" 14163.004072 \n",
" -10000.0 \n",
" -9508.0 \n",
" -9438.505655 \n",
" 0.100820 \n",
" -10285.218852 \n",
" 0.010768 \n",
" -9591.007680 \n",
" 0.000043 \n",
" 0.000118 \n",
" \n",
" \n",
" 1 \n",
" CELL.105423_1 \n",
" CELL.174282_3 \n",
" CELL.6090_2 \n",
" NeuB \n",
" GlioB \n",
" RGC \n",
" 2325.0 \n",
" 7709.0 \n",
" 14163.004072 \n",
" -9678.5 \n",
" -9508.0 \n",
" -9438.505655 \n",
" 0.031668 \n",
" -9752.127879 \n",
" 0.010768 \n",
" -9591.007680 \n",
" 0.000076 \n",
" 0.000118 \n",
" \n",
" \n",
" 2 \n",
" CELL.100795_1 \n",
" CELL.174235_3 \n",
" CELL.5906_2 \n",
" RGC \n",
" RGC \n",
" GlioB \n",
" 2893.0 \n",
" 8825.0 \n",
" 15324.653342 \n",
" -9946.5 \n",
" -9519.0 \n",
" -9458.827390 \n",
" 0.072067 \n",
" -10154.989127 \n",
" 0.009258 \n",
" -9600.700247 \n",
" 0.000055 \n",
" 0.000075 \n",
" \n",
" \n",
" 3 \n",
" CELL.105209_1 \n",
" CELL.174235_3 \n",
" CELL.5906_2 \n",
" GlioB \n",
" RGC \n",
" GlioB \n",
" 3007.0 \n",
" 8825.0 \n",
" 15324.653342 \n",
" -9693.0 \n",
" -9519.0 \n",
" -9458.827390 \n",
" 0.029907 \n",
" -9782.930904 \n",
" 0.009258 \n",
" -9600.700247 \n",
" 0.000046 \n",
" 0.000075 \n",
" \n",
" \n",
" 4 \n",
" CELL.114403_1 \n",
" CELL.174235_3 \n",
" CELL.5906_2 \n",
" RGC \n",
" RGC \n",
" GlioB \n",
" 2856.0 \n",
" 8825.0 \n",
" 15324.653342 \n",
" -8887.5 \n",
" -9519.0 \n",
" -9458.827390 \n",
" -0.105797 \n",
" -8585.344865 \n",
" 0.009258 \n",
" -9600.700247 \n",
" 0.000066 \n",
" 0.000075 \n",
" \n",
" \n",
"
\n",
"