Tutorial 2: Integrating adjacent DLPFC slices

This tutorial demonstrates SMILE’s ablility to integrate two adjacent slices (151674 and 151675). The slices are sampled from human dorsolateral prefrontal cortex (DLPFC) and the processed data can be downloaded from https://figshare.com/articles/dataset/DLPFC_slices_and_reference_scRNA-seq_data/27987548

import warnings
warnings.filterwarnings('ignore')

from stSMILE import SMILE

import scanpy as sc
import anndata as ad
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import squidpy as sq
import scipy.sparse as sp
from scipy import sparse
from scipy.sparse import csr_matrix
import math
import torch
import torch.nn as nn
import time
import torch.nn.functional as F
from itertools import chain
from scanpy import read_10x_h5
import torch.optim as optim
import sklearn
from sklearn.neighbors import kneighbors_graph
import gudhi
import networkx as nx
from torch_geometric.nn import GCNConv
import random
import os
import json 
import matplotlib.image as mpimg

Load data

section_ids = ['151674','151675']

def label_to_int(adataA, label_list, label_name):
    adata_label = np.array(adataA.obs[label_name].copy())
    for i in range(len(label_list)):
        need_index = np.where(adataA.obs[label_name]==label_list[i])[0]
        if len(need_index):
            adata_label[need_index] = i
    adataA.obs['ref'] = adata_label
    return adataA

adata_l = []
for i in range(len(section_ids)):
    adata_i = sc.read_h5ad('./dataset/DLPFC/DLPFC_'+ section_ids[i]+'_ST_final.h5ad')
    adata_i.obs_names = [x+'_'+section_ids[i] for x in adata_i.obs_names]
    adata_l.append(adata_i)

# convert label to int
label_list = ['Layer1', 'Layer2', 'Layer3', 'Layer4', 'Layer5', 'Layer6', 'WM']

for i in range(len(section_ids)):
    adata_l[i] = label_to_int(adata_l[i], label_list, 'Ground Truth')

adata0_sc = sc.read_h5ad('./dataset/DLPFC/DLPFC_sc_final.h5ad') 

adata0_sc

AnnData object with n_obs × n_vars = 19764 × 3010
    obs: 'orig.ident', 'nCount_RNA', 'nFeature_RNA', 'cell_type', 'cell_subtype', 'subject', 'condition', 'batch', 'n_genes', 'ref'
    var: 'features', 'n_cells', 'n_counts'
    uns: 'rank_genes_groups'
    obsm: 'X_pca'
    obsp: 'adj_f'

label0_list = adata0_sc.obs['cell_subtype'].tolist()
adata0_label_new = adata0_sc.obs['cell_subtype'].tolist()
for i in range(len(label0_list)):
    need_index = np.where(adata0_sc.obs['cell_subtype'] == label0_list[i])[0]    
    if len(need_index):
        for p in range(len(need_index)):
            adata0_label_new[need_index[p]] = i  

adata0_sc.obs['ref'] = pd.Series(adata0_label_new, index = adata0_sc.obs['cell_subtype'].index)
adata0_sc.obs['ref'] = adata0_sc.obs['ref'].astype(str)
adata0_sc.obs['ref'] = adata0_sc.obs['ref'].astype('category')
adata0_sc.obs['Ground Truth'] = adata0_sc.obs['cell_subtype']

adata_l.append(adata0_sc)

Run SMILE

tag_l = ['ST','ST','single cell']
in_features = len(adata_l[0].var.index)
hidden_features = 512
out_features = 50
feature_method = 'GCNConv'
alpha = 0.001
beta = 1 
lamb = 0.01 
theta = 0.001 
gamma = 10 # reconstruct 
spatial_regularization_strength= 0.9
lr=1e-3
subepochs=100
epochs=200
max_patience=50
min_stop=20
random_seed=2024
gpu=0
regularization_acceleration=True
edge_subset_sz=1000000
add_topology = True
add_feature = False
add_image = False
add_sc = True
multiscale = True
anchor_type = None
anchors_all = False
use_rep_anchor = 'embedding'
align_method = 'MMD'
anchor_size=8000
iter_comb= None
n_clusters_l = [7]
edge_weights = [1,0.1,0.1]
class_rep = 'reconstruct'

adata_l = SMILE(adata_l, tag_l, section_ids, multiscale,  n_clusters_l, in_features, feature_method, hidden_features, out_features, iter_comb, anchors_all, use_rep_anchor, alpha, beta, lamb, theta, gamma,edge_weights, add_topology, add_feature, add_image, add_sc, spatial_regularization_strength, lr=lr, subepochs=subepochs, epochs=epochs, class_rep = class_rep)

Pretraining to extract embeddings of spots...
epoch   0: train spatial C loss: 0.0000, train F loss: 38.8459,
epoch  10: train spatial C loss: 0.0000, train F loss: 30.2219,
epoch  20: train spatial C loss: 0.0000, train F loss: 24.9213,
epoch  30: train spatial C loss: 0.0000, train F loss: 21.6007,
epoch  40: train spatial C loss: 0.0000, train F loss: 19.4514,
epoch  50: train spatial C loss: 0.0000, train F loss: 17.9709,
epoch  60: train spatial C loss: 0.0000, train F loss: 16.9083,
epoch  70: train spatial C loss: 0.0000, train F loss: 16.1047,
epoch  80: train spatial C loss: 0.0000, train F loss: 15.4772,
epoch  90: train spatial C loss: 0.0000, train F loss: 14.9738,
Training classifier...
Training classifier...
torch.Size([3635, 33])
torch.Size([3566, 33])
epoch   0: overall loss: 9.7800,sc classifier loss: 3.4852,representation loss: 0.6295,within spatial regularization loss: 0.1034
epoch  10: overall loss: 3.0628,sc classifier loss: 2.3038,representation loss: 0.0759,within spatial regularization loss: 0.0733
epoch  20: overall loss: 2.3483,sc classifier loss: 1.7791,representation loss: 0.0569,within spatial regularization loss: 0.0701
epoch  30: overall loss: 1.8954,sc classifier loss: 1.4075,representation loss: 0.0488,within spatial regularization loss: 0.0831
epoch  40: overall loss: 1.6387,sc classifier loss: 1.1343,representation loss: 0.0504,within spatial regularization loss: 0.0895
epoch  50: overall loss: 1.3759,sc classifier loss: 0.9372,representation loss: 0.0439,within spatial regularization loss: 0.0971
epoch  60: overall loss: 1.2176,sc classifier loss: 0.7794,representation loss: 0.0438,within spatial regularization loss: 0.1009
epoch  70: overall loss: 1.0668,sc classifier loss: 0.6563,representation loss: 0.0410,within spatial regularization loss: 0.1038
epoch  80: overall loss: 0.9768,sc classifier loss: 0.5680,representation loss: 0.0409,within spatial regularization loss: 0.1038
epoch  90: overall loss: 0.9203,sc classifier loss: 0.4987,representation loss: 0.0422,within spatial regularization loss: 0.1036
epoch 100: overall loss: 0.9226,sc classifier loss: 0.4652,representation loss: 0.0457,within spatial regularization loss: 0.1002
epoch 110: overall loss: 0.7858,sc classifier loss: 0.4259,representation loss: 0.0360,within spatial regularization loss: 0.1017
epoch 120: overall loss: 0.7306,sc classifier loss: 0.3877,representation loss: 0.0343,within spatial regularization loss: 0.1026
epoch 130: overall loss: 0.6970,sc classifier loss: 0.3601,representation loss: 0.0337,within spatial regularization loss: 0.1042
epoch 140: overall loss: 0.6762,sc classifier loss: 0.3379,representation loss: 0.0338,within spatial regularization loss: 0.1038
epoch 150: overall loss: 0.6491,sc classifier loss: 0.3167,representation loss: 0.0332,within spatial regularization loss: 0.1039
epoch 160: overall loss: 0.7710,sc classifier loss: 0.3010,representation loss: 0.0470,within spatial regularization loss: 0.1060
epoch 170: overall loss: 0.6684,sc classifier loss: 0.2958,representation loss: 0.0373,within spatial regularization loss: 0.1028
epoch 180: overall loss: 0.6247,sc classifier loss: 0.2870,representation loss: 0.0338,within spatial regularization loss: 0.1033
epoch 190: overall loss: 0.5935,sc classifier loss: 0.2696,representation loss: 0.0324,within spatial regularization loss: 0.1037
single cell data classification: Avg Accuracy = 93.285769%


R[write to console]:                    __           __ 
   ____ ___  _____/ /_  _______/ /_
  / __ `__ \/ ___/ / / / / ___/ __/
 / / / / / / /__/ / /_/ (__  ) /_  
/_/ /_/ /_/\___/_/\__,_/____/\__/   version 6.1.1
Type 'citation("mclust")' for citing this R package in publications.



fitting ...
  |======================================================================| 100%
Identifying anchors...
Processing datasets (0, 1)
0.8404312902226964
The ratio of filtered mnn pairs: 0.8344709043272199
Aligning by anchors...
epoch 100: total loss:14.5462, train F loss: 14.5352, train C loss: 1.9671, train D loss: 0.0110
epoch 110: total loss:14.1527, train F loss: 14.1463, train C loss: 0.2483, train D loss: 0.0064
epoch 120: total loss:13.8135, train F loss: 13.8097, train C loss: 0.1978, train D loss: 0.0038
epoch 130: total loss:13.4963, train F loss: 13.4930, train C loss: 0.1789, train D loss: 0.0033
epoch 140: total loss:13.2162, train F loss: 13.2132, train C loss: 0.1657, train D loss: 0.0030
epoch 150: total loss:12.9682, train F loss: 12.9651, train C loss: 0.1568, train D loss: 0.0031
epoch 160: total loss:12.7516, train F loss: 12.7484, train C loss: 0.1465, train D loss: 0.0032
epoch 170: total loss:12.5447, train F loss: 12.5414, train C loss: 0.1370, train D loss: 0.0033
epoch 180: total loss:12.3810, train F loss: 12.3776, train C loss: 0.1306, train D loss: 0.0034
epoch 190: total loss:12.2336, train F loss: 12.2299, train C loss: 0.1184, train D loss: 0.0037
Updating classifier...
Training classifier...
epoch   0: overall loss: 15.3325,sc classifier loss: 3.6168,representation loss: 1.1716,within spatial regularization loss: 0.0920
epoch  10: overall loss: 4.0092,sc classifier loss: 2.3635,representation loss: 0.1646,within spatial regularization loss: 0.0674
epoch  20: overall loss: 2.6625,sc classifier loss: 1.7217,representation loss: 0.0941,within spatial regularization loss: 0.0751
epoch  30: overall loss: 2.1796,sc classifier loss: 1.4069,representation loss: 0.0773,within spatial regularization loss: 0.0793
epoch  40: overall loss: 1.8206,sc classifier loss: 1.1344,representation loss: 0.0686,within spatial regularization loss: 0.0856
epoch  50: overall loss: 1.7814,sc classifier loss: 0.9472,representation loss: 0.0834,within spatial regularization loss: 0.0881
epoch  60: overall loss: 1.5276,sc classifier loss: 0.8288,representation loss: 0.0699,within spatial regularization loss: 0.0913
torch.Size([3566, 33])
epoch  70: overall loss: 1.3631,sc classifier loss: 0.7348,representation loss: 0.0628,within spatial regularization loss: 0.0893
epoch  80: overall loss: 1.2525,sc classifier loss: 0.6434,representation loss: 0.0609,within spatial regularization loss: 0.0918
epoch  90: overall loss: 1.1642,sc classifier loss: 0.5708,representation loss: 0.0593,within spatial regularization loss: 0.0922
epoch 100: overall loss: 1.0934,sc classifier loss: 0.5081,representation loss: 0.0585,within spatial regularization loss: 0.0927
epoch 110: overall loss: 1.0984,sc classifier loss: 0.4562,representation loss: 0.0642,within spatial regularization loss: 0.0939
epoch 120: overall loss: 1.3328,sc classifier loss: 0.4796,representation loss: 0.0853,within spatial regularization loss: 0.0908
epoch 130: overall loss: 1.0778,sc classifier loss: 0.4287,representation loss: 0.0649,within spatial regularization loss: 0.0942
epoch 140: overall loss: 0.9877,sc classifier loss: 0.3938,representation loss: 0.0594,within spatial regularization loss: 0.0913
epoch 150: overall loss: 0.9437,sc classifier loss: 0.3653,representation loss: 0.0578,within spatial regularization loss: 0.0915
epoch 160: overall loss: 0.9127,sc classifier loss: 0.3393,representation loss: 0.0573,within spatial regularization loss: 0.0939
epoch 170: overall loss: 0.8826,sc classifier loss: 0.3186,representation loss: 0.0564,within spatial regularization loss: 0.0941
epoch 180: overall loss: 0.8607,sc classifier loss: 0.3001,representation loss: 0.0560,within spatial regularization loss: 0.0942
epoch 190: overall loss: 0.8425,sc classifier loss: 0.2846,representation loss: 0.0558,within spatial regularization loss: 0.0946
single cell data classification: Avg Accuracy = 92.724144%

adata_concat_st = ad.concat(adata_l[0:len(section_ids)], label="slice_name", keys=section_ids)

sc.tl.pca(adata_concat_st)
adata_concat_st.obsm['X_pca_old'] = adata_concat_st.obsm['X_pca'].copy()
adata_concat_st.obsm['X_pca'] = adata_concat_st.obsm['embedding'].copy()
sc.pp.neighbors(adata_concat_st)  
sc.tl.umap(adata_concat_st)

sc.tl.leiden(adata_concat_st, random_state=666, key_added="leiden", resolution=0.18)
len(list(set(adata_concat_st.obs['leiden'].tolist())))

7

Results and visualizations

from stSMILE import analysis
analysis.mclust_R(adata_concat_st, num_cluster=7, used_obsm="embedding")

fitting ...
  |======================================================================| 100%





AnnData object with n_obs × n_vars = 7201 × 3010
    obs: 'in_tissue', 'array_row', 'array_col', 'Ground Truth', 'n_genes', 'image_cluster', 'dbscan_cluster_new', 'ref', 'pd_cluster', 'slice_name', 'leiden', 'mclust'
    uns: 'pca', 'neighbors', 'umap', 'leiden'
    obsm: 'X_pca', 'features', 'features_summary_scale0.5_0.5', 'features_summary_scale0.5_1', 'features_summary_scale0.5_2', 'features_summary_scale1_0.5', 'features_summary_scale1_1', 'features_summary_scale1_2', 'features_summary_scale2_0.5', 'features_summary_scale2_1', 'features_summary_scale2_2', 'spatial', 'embedding', 'hidden_spatial', 'reconstruct', 'deconvolution', 'X_pca_old', 'X_umap'
    varm: 'PCs'
    obsp: 'distances', 'connectivities'

plt.rcParams["figure.figsize"] = (4, 4)
sc.pl.umap(adata_concat_st,color=["mclust",'Ground Truth',"slice_name"], wspace=0.4, save = 'DLPFC_umap_cluster_SMILE.pdf')  

WARNING: saving figure to file figures/umapDLPFC_umap_cluster_SMILE.pdf

Results and visualizations

import matplotlib.pyplot as plt
from sklearn.metrics import adjusted_rand_score as ari_score
from sklearn.metrics import normalized_mutual_info_score as nmi_score
# split to each data
Batch_list = []
for section_id in section_ids:
    Batch_list.append(adata_concat_st[adata_concat_st.obs['slice_name'] == section_id])

spot_size = 200
title_size = 12
ARI_list = []
NMI_list = []
for bb in range(len(section_ids)):
    ARI_list.append(round(ari_score(Batch_list[bb].obs['Ground Truth'], Batch_list[bb].obs['mclust']), 2))
    NMI_list.append(round(nmi_score(Batch_list[bb].obs['Ground Truth'], Batch_list[bb].obs['mclust']), 2))

fig, ax = plt.subplots(2, 1, figsize=(3.5, 7), gridspec_kw={'wspace': 0.05, 'hspace': 0.1})
_sc_0 = sc.pl.spatial(Batch_list[0], img_key=None, color=['mclust'], title=[''],
                      legend_loc=None, legend_fontsize=12, show=False, ax=ax[0], frameon=False,
                      spot_size=spot_size)
_sc_0[0].set_title("ARI=" + str(ARI_list[0])+",NMI=" + str(NMI_list[0]), size=title_size)
_sc_1 = sc.pl.spatial(Batch_list[1], img_key=None, color=['mclust'], title=[''],
                      legend_loc=None, legend_fontsize=12, show=False, ax=ax[1], frameon=False,
                      spot_size=spot_size)
_sc_1[0].set_title("ARI=" + str(ARI_list[1])+",NMI=" + str(NMI_list[1]), size=title_size)
plt.savefig("DLPFC_spatial_SMILE.pdf") 
plt.show()

adata_l[0].uns['deconvolution']

	AAACAAGTATCTCCCA-1_151674_151674	AAACAATCTACTAGCA-1_151674_151674	AAACACCAATAACTGC-1_151674_151674	AAACAGAGCGACTCCT-1_151674_151674	AAACAGCTTTCAGAAG-1_151674_151674	AAACAGGGTCTATATT-1_151674_151674	AAACAGTGTTCCTGGG-1_151674_151674	AAACATTTCCCGGATT-1_151674_151674	AAACCCGAACGAAATC-1_151674_151674	AAACCGGGTAGGTACC-1_151674_151674	...	TTGTGTATGCCACCAA-1_151674_151674	TTGTGTTTCCCGAAAG-1_151674_151674	TTGTTAGCAAATTCGA-1_151674_151674	TTGTTCAGTGTGCTAC-1_151674_151674	TTGTTGTGTGTCAAGA-1_151674_151674	TTGTTTCACATCCAGG-1_151674_151674	TTGTTTCATTAGTCTA-1_151674_151674	TTGTTTCCATACAACT-1_151674_151674	TTGTTTGTATTACACG-1_151674_151674	TTGTTTGTGTAAATTC-1_151674_151674
Ex_3_L4_5	6.969129e-03	6.174900e-04	1.027352e-08	2.907331e-02	1.151733e-01	2.179292e-01	6.726495e-05	1.173180e-02	3.489990e-03	6.216561e-01	...	1.107924e-02	3.292672e-03	1.327825e-01	3.654320e-01	7.285582e-01	2.056747e-07	7.804774e-08	1.532008e-02	6.899436e-05	3.995313e-06
Mix_2	1.144664e-07	3.622714e-07	1.838581e-06	7.544602e-08	3.360677e-01	1.030835e-03	1.469804e-04	9.631843e-07	9.051941e-10	6.517683e-04	...	2.440290e-04	4.984937e-04	4.404069e-06	3.596538e-03	1.487128e-03	2.864074e-05	9.462240e-06	1.683173e-02	6.111187e-05	2.503635e-09
Astros_3	7.542109e-13	4.976497e-09	3.281698e-05	5.876007e-13	1.142282e-05	2.532608e-05	1.662139e-04	5.499309e-12	1.104752e-12	7.876854e-08	...	1.222819e-04	8.953258e-04	9.232067e-12	1.188053e-08	1.772218e-08	2.183260e-03	3.123484e-04	6.731632e-03	1.586484e-05	9.812313e-09
Inhib_5	2.599940e-05	3.712677e-06	4.739135e-07	2.806954e-06	1.215782e-05	2.259663e-07	1.252311e-05	1.007999e-03	2.926604e-09	2.613846e-09	...	3.426004e-07	2.554535e-06	1.385688e-04	9.882564e-07	2.053443e-07	3.610777e-06	2.119538e-06	4.931921e-05	3.495906e-06	5.199091e-08
Oligos_2	3.322113e-10	6.356483e-09	3.128458e-01	1.226389e-10	3.514589e-06	1.537011e-06	2.353112e-03	1.832942e-09	9.932827e-12	1.220030e-08	...	8.208621e-06	1.067566e-04	1.000387e-09	3.041905e-08	2.118467e-08	4.344048e-01	3.445954e-01	9.848450e-04	3.311624e-04	8.219440e-10
OPCs_1	2.303407e-08	6.548941e-07	1.453799e-05	1.375264e-08	5.889276e-06	3.458549e-05	1.181179e-04	1.105657e-08	2.375208e-06	9.104296e-07	...	2.467473e-04	1.040747e-03	6.711546e-09	7.872727e-07	2.454133e-06	1.097253e-03	6.136653e-05	3.437947e-03	2.492627e-05	2.216885e-03
Mix_5	7.022922e-02	2.340805e-05	2.529608e-05	5.073861e-02	4.253091e-03	1.422330e-04	1.955855e-04	3.317325e-02	1.347937e-03	3.805917e-05	...	3.302010e-05	1.225939e-04	1.175446e-03	1.850879e-02	2.356324e-02	1.551443e-04	6.335450e-05	7.949749e-04	6.539816e-05	3.841993e-05
Inhib_6_SST	3.809572e-05	7.408547e-11	4.430857e-07	1.531785e-05	1.065465e-05	4.851082e-08	8.072549e-07	2.037239e-05	5.269946e-08	2.654861e-09	...	7.926588e-09	5.665231e-08	2.914060e-05	1.695641e-02	1.828895e-02	4.030802e-06	1.114944e-06	2.130349e-06	2.774746e-08	8.693223e-10
Oligos_1	4.715230e-10	1.093477e-08	6.251897e-01	2.400796e-10	3.440437e-06	9.618647e-06	1.489260e-01	1.626619e-09	3.958604e-11	9.314818e-08	...	8.908093e-05	6.741536e-04	2.843505e-10	3.177034e-09	3.359012e-09	4.213376e-01	5.292145e-01	1.707045e-03	9.126685e-02	4.159424e-10
Mix_1	9.387812e-09	2.771037e-09	8.217115e-07	6.523106e-09	1.420885e-01	8.507208e-04	5.206808e-04	5.062454e-08	4.212891e-11	1.321318e-04	...	9.792802e-05	3.395937e-04	2.503545e-08	2.687180e-05	2.250277e-05	3.613956e-06	4.228411e-06	5.304500e-03	3.774091e-04	6.675904e-12
Inhib_3_SST	3.131800e-08	7.564521e-03	2.282922e-08	1.990833e-08	9.484770e-07	2.559283e-06	2.982678e-06	1.011834e-07	2.429055e-07	6.606045e-08	...	1.448983e-05	2.983885e-05	1.183360e-08	2.508879e-08	2.951299e-08	1.085739e-07	7.110510e-08	1.427447e-04	8.025528e-07	3.331172e-02
Ex_6_L4_6	1.106280e-07	1.275956e-06	2.031759e-08	7.696657e-08	2.844928e-04	3.237899e-05	1.330352e-04	2.941007e-06	4.668109e-09	9.684834e-06	...	1.194815e-04	2.546169e-04	8.810264e-03	2.770237e-03	3.959417e-03	5.334537e-07	2.031766e-07	1.943589e-03	1.828023e-04	2.783026e-09
Ex_7_L4_6	4.053102e-09	7.869427e-10	5.796662e-07	2.719557e-09	1.915012e-03	8.098045e-04	5.627084e-03	4.114865e-09	1.764348e-09	4.648007e-05	...	9.406718e-04	9.037836e-03	7.309504e-09	3.190497e-03	4.209323e-02	1.141165e-05	9.423859e-06	4.144669e-03	5.858894e-03	3.305049e-10
Ex_4_L_6	3.172382e-07	6.770508e-07	8.324937e-08	6.476919e-07	1.925441e-01	2.307862e-01	6.572544e-05	3.285137e-07	3.514043e-07	2.850848e-01	...	1.186501e-02	5.878960e-03	1.668930e-07	9.200487e-04	4.525761e-03	1.131073e-06	4.306189e-07	4.911749e-02	2.096841e-05	5.733975e-08
Mix_4	5.353738e-06	6.300014e-05	4.285703e-06	6.457186e-06	2.401002e-02	3.462924e-04	1.090475e-05	1.166674e-05	6.393877e-07	5.348081e-04	...	6.903738e-05	8.091881e-05	5.745594e-06	1.284958e-03	1.178578e-03	4.821973e-05	8.093351e-06	3.979215e-03	1.448414e-06	4.277290e-05
Inhib_2_VIP	4.527714e-06	1.354578e-05	9.197858e-08	1.842564e-06	1.427521e-06	1.365741e-06	1.066277e-06	7.445340e-05	9.446614e-08	4.357029e-08	...	2.420397e-06	5.483661e-06	1.623088e-04	2.007017e-06	1.137843e-06	3.662107e-07	1.509226e-07	8.139470e-05	9.484784e-08	6.822607e-05
Astros_2	1.761832e-06	6.635826e-05	3.024704e-05	1.313114e-06	8.566742e-05	2.368000e-06	2.970858e-05	2.334632e-06	2.129988e-06	8.542484e-08	...	6.076156e-06	2.831762e-05	2.195614e-07	4.711030e-06	2.515737e-06	1.217556e-03	1.001168e-04	8.825109e-04	3.072990e-06	1.842220e-03
Astros_1	1.201523e-12	1.953085e-09	1.224096e-03	6.766249e-13	1.439081e-04	5.914796e-05	1.077180e-03	1.619942e-11	5.432529e-14	3.611861e-07	...	1.348757e-04	8.616096e-04	5.370073e-12	4.659470e-09	3.766855e-09	4.241443e-03	3.825556e-03	9.002970e-03	1.639025e-04	9.392908e-11
Ex_2_L5	5.746522e-06	4.701679e-09	6.476725e-08	5.760590e-06	1.292586e-02	1.810356e-04	1.060314e-05	3.284260e-05	1.765196e-09	1.468756e-04	...	1.546418e-06	3.254646e-06	5.624586e-06	1.273811e-04	1.186338e-04	2.141313e-07	1.801047e-07	4.797298e-05	3.701893e-06	4.880471e-12
OPCs_2	1.126917e-13	4.612336e-11	1.095641e-04	2.999705e-14	1.826984e-08	7.478603e-08	1.748297e-05	8.929763e-13	1.642399e-13	1.425015e-10	...	1.583036e-06	2.288817e-05	8.515883e-13	1.005902e-10	1.236756e-10	1.836568e-02	5.616478e-04	2.700382e-04	1.055998e-06	1.909537e-08
Ex_8_L5_6	5.428657e-10	5.606627e-08	2.110268e-04	5.391343e-10	2.839633e-02	3.337331e-01	2.525081e-01	2.868516e-09	1.863032e-10	7.810751e-02	...	3.863671e-01	4.729090e-01	3.613515e-09	4.788989e-06	1.689251e-05	3.078576e-03	1.293664e-03	4.766280e-01	4.005852e-01	2.234830e-09
Inhib_1	1.470631e-06	3.972175e-07	3.982941e-06	1.133233e-06	2.783863e-06	3.230669e-09	6.162509e-09	1.660090e-06	2.271773e-08	6.102157e-10	...	3.125852e-10	1.509004e-09	4.700690e-07	4.685031e-04	2.209392e-04	1.621055e-05	2.386047e-06	4.400691e-07	4.863112e-11	8.454985e-07
Ex_9_L5_6	4.922706e-06	4.455678e-07	1.986375e-08	7.068288e-06	2.335303e-04	1.414419e-04	7.026473e-05	2.542575e-06	7.744870e-06	3.626575e-05	...	2.523793e-05	4.216139e-05	4.067813e-07	1.640185e-04	5.512940e-04	2.137241e-07	1.032411e-07	1.253849e-04	3.691699e-05	7.113815e-08
Inhib_8_PVALB	3.969033e-04	5.369036e-01	2.649109e-08	2.762289e-04	2.431192e-04	8.852249e-05	3.312487e-05	2.703326e-03	1.099732e-04	6.824694e-06	...	1.061196e-04	1.624674e-04	5.408885e-04	2.668971e-05	1.144796e-05	2.943015e-07	1.686579e-07	8.204745e-04	3.351401e-05	9.823893e-02
Inhib_4_SST	5.559318e-08	8.254041e-07	1.008312e-08	2.059556e-08	1.948467e-05	1.998858e-07	9.184345e-07	5.318890e-07	2.553563e-10	4.052869e-09	...	7.159372e-08	3.145651e-07	2.685139e-07	1.506060e-05	5.135757e-06	4.504801e-08	4.374684e-08	1.086342e-05	1.489151e-07	5.204655e-09
Inhib_7_PVALB	3.771151e-03	7.065056e-06	1.021373e-08	2.133727e-03	1.390819e-02	5.751749e-05	5.128600e-06	5.476939e-02	1.131040e-06	1.483429e-05	...	4.211597e-06	8.107341e-06	4.601649e-01	4.292178e-01	1.164777e-01	1.293240e-07	6.188822e-08	1.499547e-04	2.128651e-06	2.483982e-08
Mix_3	2.188812e-01	1.726181e-01	5.891335e-08	2.591775e-01	5.028176e-03	2.073113e-04	6.476233e-05	2.764076e-01	2.085457e-01	1.972519e-05	...	1.438308e-04	2.416857e-04	1.978173e-01	1.488124e-01	4.449908e-02	8.786663e-07	3.849669e-07	9.569440e-04	7.206837e-05	1.072768e-01
Ex_1_L5_6	2.024828e-12	1.926577e-11	1.444598e-05	9.708949e-13	7.182638e-02	9.340792e-04	2.515586e-04	1.367040e-11	7.509047e-14	2.161433e-05	...	1.625330e-03	7.510009e-03	3.580937e-10	1.984929e-03	5.716261e-03	2.537716e-03	1.905493e-04	5.232777e-02	4.012349e-05	5.512795e-12
Micro/Macro	1.864344e-06	1.119673e-04	6.629099e-04	1.160585e-06	3.573515e-04	4.381800e-05	7.142534e-04	1.913180e-05	1.023166e-08	1.224546e-05	...	4.084556e-05	4.492511e-05	1.863973e-05	4.098184e-07	1.267596e-07	1.708352e-04	3.516524e-04	2.055096e-03	1.912791e-04	1.434237e-07
Oligos_3	3.709469e-08	7.122149e-07	1.181898e-03	2.290434e-08	5.436231e-05	1.418966e-05	8.566926e-03	7.681808e-08	1.285053e-07	2.595675e-07	...	2.037578e-04	1.208978e-03	7.327127e-07	4.243827e-05	4.977611e-05	9.372629e-02	9.014801e-03	4.973021e-03	4.162602e-03	4.830601e-06
Endo	3.435871e-08	7.732856e-05	5.844417e-02	2.636470e-08	3.998118e-03	3.076089e-02	5.755958e-01	2.106401e-07	2.409081e-09	1.783860e-03	...	1.161141e-01	3.119579e-01	8.286518e-08	4.794928e-07	8.316435e-07	1.735740e-02	1.103736e-01	2.765187e-01	4.892079e-01	1.000246e-07
Ex_10_L2_4	6.994792e-01	2.819237e-01	6.619200e-07	6.581998e-01	3.837510e-02	1.812108e-01	2.703726e-03	6.199418e-01	7.864754e-01	9.605638e-03	...	4.702660e-01	1.827232e-01	1.982044e-01	1.340843e-03	1.113589e-03	6.538758e-06	2.657908e-06	6.432272e-02	7.215655e-03	7.569538e-01
Ex_5_L5	1.827829e-04	7.442681e-07	5.615697e-09	3.570721e-04	8.015924e-03	5.623616e-04	2.440965e-06	9.460215e-05	1.593965e-05	2.088912e-03	...	2.733703e-05	1.451911e-05	1.373379e-04	5.099257e-03	7.534959e-03	1.201819e-07	2.699033e-08	3.338726e-04	4.855864e-07	5.201670e-08

33 rows × 3635 columns

# write out the deconvolution result
for i in range(len(adata_l)-1):
    adata_i = adata_l[i].copy()
    dev_i = adata_i.uns['deconvolution']
    dev_i.to_csv('DLPFC_SMILE_dev_'+ section_ids[i]+'.csv', sep='\t')