stereo.core.StPipeline.find_marker_genes in multi-sample

[sparkumr]

stereo.core.StPipeline.find_marker_genes： The set control group cannot be identified in the multi-sample, and when the program is running, it will only compare the difference between the case group and all the other groups

[tanliwei-coder]

I don't clear how you run find_marker_genes with multi-sample, could you show me your codes?

[sparkumr]

import sys import os from natsort import natsorted import stereo as st from stereo.core.ms_data import MSData from stereo.core.ms_pipeline import slice_generator import matplotlib.pyplot as plt import dill import warnings from stereo.algorithm.co_occurrence import CoOccurrence from stereo.core.ms_pipeline import slice_generator import pandas as pd import numpy as np warnings.filterwarnings('ignore')

    data_dir = '/public/home/sangchl/work_result/Spatial_Transcriptomics/register/3/intergrate_sc_spa/data'

    data_list=[]
    for fn in os.listdir(data_dir):
            data_list.append(os.path.join(data_dir, fn))

    data_list = natsorted(data_list)

    ms_data = MSData(_relationship='other', _var_type='intersect')

    for sample in data_list:
            ms_data += st.io.read_gef(file_path=sample, bin_type='cell_bins')

    ms_data.names=['Control','Ccnk']
    ms_data.integrate()
    ms_data.tl.cal_qc(scope=slice_generator[:],mode='integrate')

    ms_data.plt.spatial_scatter(
            scope=slice_generator[:],
            mode='integrate',
            plotting_scale_width=10,          # the width of scale
            reorganize_coordinate=4,          # the number of plots in each row
            horizontal_offset_additional=20,  # adjustment for horizontal distance
            vertical_offset_additional=20     # adjustment for vertical distance
            )
    ms_data.tl.filter_cells(min_gene=20, min_n_genes_by_counts=3, pct_counts_mt=5, scope=slice_generator[:], mode='integrate', inplace=True)
    ms_data.tl.raw_checkpoint()
    ms_data.tl.normalize_total(target_sum=10000)
    ms_data.tl.log1p()
    ms_data.tl.highly_variable_genes(
            min_mean=0.0125,
            max_mean=3,
            min_disp=0.5,
            n_top_genes=2000,
            res_key='highly_variable_genes',
            scope=slice_generator[:],
            mode='integrate'
            )
    ms_data.plt.highly_variable_genes(res_key='highly_variable_genes')
    ms_data.tl.scale(max_value=10, zero_center=True)

    ms_data.tl.pca(
            use_highly_genes=False,
            n_pcs=30,
            res_key='pca',
            scope=slice_generator[:],
            mode='integrate'
            )
    ms_data.tl.neighbors(
            pca_res_key='pca',
            n_pcs=30,
            res_key='neighbors',
            scope=slice_generator[:],
            mode='integrate'
            )

    ms_data.tl.umap(
            pca_res_key='pca',
            neighbors_res_key='neighbors',
            res_key='umap',
            scope=slice_generator[:],
            mode='integrate'
            )

    ms_data.tl.louvain(
            neighbors_res_key='neighbors',
            res_key='louvain',
            scope=slice_generator[:],
            mode='integrate'
            )

    annotation_dict = {
            '1':'VLMC',
            '2':'VLMC',
            '3':'VLMC',
            '4':'Immature Neurons',
            '5':'Red Blood Cell',
            '6':'Apical Progenitors 1',
            '7':'Apical Progenitors 2',
            '8':'Apical Progenitors 3',
            '9':'Apical Progenitors 4',
            '10':'Medial Forebrain Progenitors 1',
            '11':'unknown',
            '12':'VLMC',
            '13':'VLMC',
            '14':'Medial Forebrain Progenitors 2',
            '15':'Microglia'
            }

    ms_data.tl.annotation(
            annotation_information=annotation_dict,
            cluster_res_key='louvain',
            res_key='anno_louvain'
            )

    ms_data.tl.find_marker_genes(cluster_res_key='anno_louvain',
                                    output='/public/home/sangchl/work_result/Spatial_Transcriptomics/register/3/intergrate_sc_spa/'+ case + 'vs' + control + ".csv", 
                                    use_highly_genes=False,
                                    use_raw =True,
                                    case_groups=case,
                                    control_groups=control)

[sparkumr]

The case group and the control group here are the cell subsets I commented, and it can be run, but it doesn't seem to be used for the control group information

[tanliwei-coder]

In your case, the case_groups and control_groups must be a group or a list of groups in the anno_louvain, I don't see that what values you set into them, and, how did you judge the control group didn't be used?

[sparkumr]

在你的例子中，case_groups和control_groups必须是 中的组或组列表，我看不出你在其中设置了什么值，而且，你如何判断对照组没有被使用？anno_louvain

The case and control here are created by myself according to the format in the ms_data, and the control group is not applied because I tried different control values, but he both returned the same result