CellphoneDB的分析流程(方法二)学习

# 展示数据库版本
from IPython.display import HTML, display
from cellphonedb.utils import db_releases_utils

display(HTML(db_releases_utils.get_remote_database_versions_html()['db_releases_html_table']))

# 设定数据库版本
cpdb_version = 'v5.0.0'

# 数据库资料所需的存放文件的路径
# 路径自行设定
cpdb_target_dir = os.path.join('/home/jovyan/cpdb_tutorial/db/test', cpdb_version)

# 开始下载数据
from cellphonedb.utils import db_utils

db_utils.download_database(cpdb_target_dir, cpdb_version)

conda create -n cpdb python=3.8
conda activate cpdb

# 注册内核
pip install ipykernel
python -m ipykernel install --user --name cpdb --display-name "Python(cpdb)"

# 终端中安装
pip install cellphonedb

# 需要在终端中安装ktplotspy，这是一个适用于cellphoneDB的可视化工具
pip install ktplotspy

# 有一些包是多余的，但是也可以导入
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import scanpy as sc
import numpy as np
import anndata as ad
import ktplotspy as kpy
import os

# 确认位置
os.getcwd()

pd.set_option('display.max_columns', 100) # 设置pandas显示的最大列数为100，如果超过默认显示的数量，pandas会省略一些列显示。
os.chdir('./3-CellphoneDB/') # 移动到相应的文件夹中

# 读取adata数据
adata = sc.read_h5ad("../scRNA_V5.h5ad")

# 归一化数据
sc.pp.normalize_per_cell(adata, counts_per_cell_after=1e4)

# 保存一个新的数据
adata.write('./adata_Nor_counts.h5ad')

# 获取adata的细胞注释文件
import pandas as pd

# 创建一个包含Cell和Cell_type两列的DataFrame
df = pd.DataFrame({
    'Cell': adata.obs_names,
    'Cell_type': adata.obs['celltype']
})

# 显示前几行（可选）
print(df.head())

# 如果需要保存为 CSV 文件：
df.to_csv('cell_annotations.csv', index=False)

cpdb_file_path = '../cpdb_database/v5/cellphonedb.zip'
meta_file_path = './cell_annotations.csv'
counts_file_path = './adata_Nor_counts.h5ad'
#microenvs_file_path = 'cell_annotations.csv'
#active_tf_path = 'data/active_TFs.tsv'
out_path = './'

from cellphonedb.src.core.methods import cpdb_statistical_analysis_method

cpdb_results = cpdb_statistical_analysis_method.call(
    cpdb_file_path = cpdb_file_path,                 # 必填: CellphoneDB database zip file.
    meta_file_path = meta_file_path,                 # 必填: tsv file defining barcodes to cell label.
    counts_file_path = counts_file_path,             # 必填: normalized count matrix - a path to the counts file, or an in-memory AnnData object
    counts_data = 'hgnc_symbol',                     # 在计数矩阵中定义基因注释信息.
    #active_tfs_file_path = active_tf_path,           # 可选: 定义细胞类型及其活跃的转录因子（TFs）。
    #microenvs_file_path = microenvs_file_path,       # 可选 (default: None): 定义了每个微环境中的细胞数量.
    score_interactions = True,                       # 可选: 是否要记录互动情况. 
    iterations = 1000,                               # 表示在分析中执行的随机置换次数。.
    threshold = 0.1,                                 # 定义了在分析中使用某个基因所需的最小细胞表达百分比。.
    threads = 5,                                     # 分析中使用的线程数量。.
    debug_seed = 42,                                 # 调试随机种子。要禁用，请设置为 >= 0。.
    result_precision = 3,                            # Sets the rounding for the mean values in significan_means.
    pvalue = 0.05,                                   # P值的阈值
    subsampling = False,                             # 启用数据的子抽样（几何草图）。
    subsampling_log = False,                         # （必填）为未经过对数转换的数据输入启用 subsampling log1p。
    subsampling_num_pc = 100,                        # 通过几何草图进行子抽样的组件数量（默认：100）。
    subsampling_num_cells = 1000,                    # 要进行子抽样的细胞数量（整数）（默认：数据集的 1/3）。
    separator = '|',                                 # 设置用于分隔结果数据框中细胞的字符串 "cellA|CellB"。
    debug = False,                                   # 将分析过程中使用的所有中间表以 pkl 格式保存。
    output_path = out_path,                          # Path to save results.
    output_suffix = None                             # Replaces the timestamp in the output files by a user defined string in the  (default: None).
    )

list(cpdb_results.keys())

# ['deconvoluted',
#  'deconvoluted_percents',
#  'means',
#  'pvalues',
#  'significant_means',
#  'CellSign_active_interactions',
#  'CellSign_active_interactions_deconvoluted',
#  'interaction_scores']

kpy.plot_cpdb_heatmap(pvals = cpdb_results['pvalues'],
                      degs_analysis = False,
                      figsize = (5, 5),
                      title = "Sum of significant interactions")

kpy.plot_cpdb(
    adata=adata,
    cell_type1="B cell",#比如这里我们需要看B cells细胞与其他细胞的通讯
    cell_type2=".",  # 表示所有细胞
    means=cpdb_results['means'],
    pvals=cpdb_results['pvalues'],
    celltype_key="celltype",
    gene_family="chemokines",
    figsize=(8, 5),
    title="interacting interactions",
    max_size = 3,
    highlight_size = 0.75
)

kpy.plot_cpdb(
    adata = adata,
    cell_type1 = "B cell",
    cell_type2 = ".",
    means = cpdb_results['means'],
    pvals = cpdb_results['pvalues'],
    celltype_key = "celltype",
    genes = ["TGFB2", "CSF1R"],
    figsize = (10, 3),
    title = "Interactions between\nB cell and other",
    max_size = 3,
    highlight_size = 0.75,
    degs_analysis = False,
    standard_scale = True,
    interaction_scores = cpdb_results['interaction_scores'],
    scale_alpha_by_interaction_scores = True
)