OSCR

Region- and cell type-specific changes in gene expression in the cerebellum after classical fear conditioning.

Code ↔ Paper

17 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 17 matches
  1. [1] § Methods › Visium spatial transcriptomics analysis › Differential gene expression analysis ↔ Codes/Analysis/spatial_pseudobulk_deseq.R, lines 274–347 · score 0.75 · DESeq, independentFiltering, pairwise comparison, LRT DEGs, alpha, HC
  2. [2] § Methods › snRNA-seq data analysis › Preprocess and annotation ↔ Codes/Analysis/preprocess_annotate_snrna.ipynb, lines 599–615 · score 0.72 · Seurat v3, sc.pp.highly_variable_genes, flavor, latent, scVI, batch
  3. [3] § Methods › Visium spatial transcriptomics analysis › Spatial transcriptomics mapping and preprocess ↔ Codes/Analysis/preprocess_annotate_snrna.ipynb, lines 599–615 · score 0.72 · Seurat v3, sc.pp.highly_variable_genes, flavor, latent, scVI, batch
  4. [4] § Methods › Visium spatial transcriptomics analysis › Annotation of spatial clusters ↔ Codes/Analysis/annotation_cell2location.ipynb, lines 277–288 · score 0.71 · detection alpha, cell2location, cell abundance, training, model
  5. [5] § Results › Specific inhibitory neuron subtype expressing Kit is highly associated with fear conditioning ↔ Codes/Analysis/preprocess_annotate_snrna.ipynb, lines 794–801 · score 0.69 · Inh Piezo2, Inh Zfhx4, inhibitory neurons, Inh Kit, snRNA, cells
  6. [6] § Results › snRNA-seq in the DCN reveals cell type- and learning phase-specific transcriptional changes ↔ Codes/Figure/Figure5.R, lines 1–42 · score 0.66 · snRNA, CD DEGs, LRT DEGs, UMAP, astrocyte, Quadrant
  7. [7] § Methods › snRNA-seq data analysis › Transcription factor activity analysis ↔ Codes/Analysis/pySCENIC_downstream.R, lines 1–55 · score 0.60 · pySCENIC, motifs, AUC, adjacency, mm10, scores
  8. [8] § Methods › Visium spatial transcriptomics analysis › Spatial transcriptomics mapping and preprocess ↔ Codes/Analysis/spatialclustering_squidpy.ipynb, lines 55–59 · score 0.60 · joint graph, Squidpy, space, resolution, Leiden, sc
  9. [9] § Results › Specific inhibitory neuron subtype expressing Kit is highly associated with fear conditioning ↔ Codes/Analysis/snrna_MAST.R, lines 1584–1642 · score 0.59 · Inh Piezo2, Inh Zfhx4, Inh Kit, snRNA, DEGs, DCN
  10. [10] § Methods › snRNA-seq data analysis › Differential gene expression analysis ↔ Codes/Analysis/snrna_MAST.R, lines 928–996 · score 0.59 · pairwise comparison, log2FC, MAST, zlm, Hurdle, LRT
  11. [11] § Results › Specific inhibitory neuron subtype expressing Kit is highly associated with fear conditioning ↔ Codes/Analysis/preprocess_annotate_snrna.ipynb, lines 794–801 · score 0.57 · Inh Zfhx4, inhibitory neurons, Inh Kit, neuronal, cells
  12. [12] § Results › Specific inhibitory neuron subtype expressing Kit is highly associated with fear conditioning ↔ Codes/Figure/Figure6.R, lines 1–41 · score 0.56 · inhibitory neuron, Inh Kit, UMAP, Quadrant, Zfhx4, Grm5
  13. [13] § Results › Spatial transcriptomic analysis in the cerebellum after classical fear conditioning ↔ Codes/Figure/Figure2.R, lines 43–111 · score 0.56 · granular layer, molecular layer, ventricle, medulla, Purkinje, seq
  14. [14] § Results › Spatial transcriptomic analysis in the cerebellum after classical fear conditioning ↔ Codes/Analysis/spatial_pseudobulk_deseq.R, lines 274–347 · score 0.55 · granular layer, molecular layer, ventricle, medulla, Purkinje, seq
  15. [15] § Methods › snRNA-seq data analysis › Transcription factor activity analysis ↔ Codes/Analysis/pySCENIC_downstream.R, lines 160–207 · score 0.54 · regulon activity, Bonferroni, RAS, scores, TF, cell
  16. [16] § Results › Specific inhibitory neuron subtype expressing Kit is highly associated with fear conditioning ↔ Codes/Figure/Figure6.R, lines 1–41 · score 0.54 · inhibitory neurons, Inh Kit, Zfhx4, Grm5, regulator, neuronal
  17. [17] § Methods › Visium spatial transcriptomics analysis › Differential gene expression analysis ↔ Codes/Analysis/spatial_pseudobulk_deseq.R, lines 65–84 · score 0.53 · mitochondrial genes, DESeq2, summing, subsets, cell

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Jupyter notebook · 970 lines · 31 KB · CC-BY-4.0 · 4 matches

  1. # %%
  2. import gc
  3. import pandas as pd
  4. import numpy as np
  5. import scanpy as sc
  6. import anndata as ad
  7. import scvi
  8. import torch
  9. import anndata
  10. import copy
  12. from rich import print
  13. from scib_metrics.benchmark import Benchmarker
  14. from scvi.model.utils import mde
  15. from scvi_colab import install
  17. #import scrublet as scr
  19. import matplotlib.pyplot as plt
  20. import random
  21. import matplotlib as mpl
  22. import seaborn as sns
  24. # Set font
  25. mpl.rcParams['pdf.fonttype'] = 42
  26. mpl.rcParams['font.family'] = ['Arial']
  28. # Check torch
  29. torch.cuda.is_available()
  31. # %%
  32. scvi.settings.seed = 2023
  34. # %%
  35. import os
  36. os.chdir('/data1/Spatial_DCN/')
  38. # %%
  39. import gc
  41. # %%
  42. # set seed for randomness
  43. torch.manual_seed(2023)
  44. random.seed(2023)
  45. np.random.seed(2023)
  46. torch.backends.cudnn.deterministic = True
  47. torch.backends.cudnn.benchmark = False
  49. # %%
  50. # plot settings
  51. title_fs = 16
  52. axis_label_fs = 14
  53. axis_tick_fs = 12
  54. legend_fs = 10
  56. # %% [markdown]
  57. # # Import raw data
  59. # %%
  60. adata_hc = sc.read_h5ad('Resources/snRNA_samples/HC/HC_matrix/HC_matrix.h5ad')
  61. adata_cd = sc.read_h5ad('Resources/snRNA_samples/FC/FC_matrix/FC_matrix.h5ad')
  62. adata_tn = sc.read_h5ad('Resources/snRNA_samples/TN/TN_matrix/TN_matrix.h5ad')
  64. # %%
  65. # set same name as Visium
  66. adata_cd.obs['sample'] = 'CD'
  68. # %%
  69. adata_lst = [adata_hc, adata_cd, adata_tn]
  71. # %%
  72. samp_lst = ['HC', 'CD', 'TN']
  74. # %%
  75. for ad in adata_lst:
  76. ad.var_names = ad.var['gene_symbols'].astype(str)
  78. # %%
  79. for adata in adata_lst:
  80. adata.var_names_make_unique()
  82. # %%
  83. for i in range(3):
  84. print(f"Number of spots/genes for {samp_lst[i]} sample: {len(adata_lst[i].obs)}/{len(adata_lst[i].var)}")
  86. # %% [markdown]
  87. # # Doublet detection
  89. # %%
  90. hc_cnt = pd.DataFrame(data=adata_hc.X.toarray(), index=adata_hc.obs_names, columns=adata_hc.var_names)
  91. cd_cnt = pd.DataFrame(data=adata_cd.X.toarray(), index=adata_cd.obs_names, columns=adata_cd.var_names)
  92. tn_cnt = pd.DataFrame(data=adata_tn.X.toarray(), index=adata_tn.obs_names, columns=adata_tn.var_names)
  94. # %%
  95. hc_scrub = scr.Scrublet(hc_cnt, expected_doublet_rate= 0.008 * 7.6)
  96. cd_scrub = scr.Scrublet(cd_cnt, expected_doublet_rate= 0.008 * 10.8)
  97. tn_scrub = scr.Scrublet(tn_cnt, expected_doublet_rate= 0.008 * 7.9)
  99. # %%
  100. hc_db_scr, hc_db_pred = hc_scrub.scrub_doublets()
  101. cd_db_scr, cd_db_pred = cd_scrub.scrub_doublets()
  102. tn_db_scr, tn_db_pred = tn_scrub.scrub_doublets()
  104. # %%
  105. db_scr_lst = [hc_db_scr, cd_db_scr, tn_db_scr]
  106. db_pred_lst = [hc_db_pred, cd_db_pred, tn_db_pred]
  108. # %%
  109. for i in range(3):
  110. adata_lst[i].obs['doublet_score'] = db_scr_lst[i]
  111. adata_lst[i].obs['doublet_prediction'] = db_pred_lst[i]
  113. # %%
  114. for adata in adata_lst:
  115. print(len(adata[adata.obs['doublet_prediction'] == True].obs_names))
  117. # %%
  118. for i in range(3):
  119. adata_lst[i] = adata_lst[i][adata_lst[i].obs['doublet_prediction'] == False]
  120. print(f"Number of spots/genes for {samp_lst[i]} sample: {len(adata_lst[i].obs)}/{len(adata_lst[i].var)}")
  122. # %% [markdown]
  123. # # QC
  125. # %%
  126. # Basic filtering
  127. for adata in adata_lst:
  128. sc.pp.filter_genes(adata, min_cells= 3)
  130. for i in range(3):
  131. print(f"Number of spots/genes for {samp_lst[i]} sample: {len(adata_lst[i].obs)}/{len(adata_lst[i].var)}")
  133. # %%
  134. for adata in adata_lst:
  135. adata.var['mt'] = adata.var_names.str.startswith('mt-')
  136. sc.pp.calculate_qc_metrics(adata, qc_vars=['mt'], percent_top=None, log1p=False, inplace=True)
  138. # %%
  139. qc_lst = ['total_counts', 'n_genes_by_counts', 'pct_counts_mt']
  141. # %%
  142. concat_data = sc.concat(
  143. adata_lst,
  144. label="sample",
  145. keys=['HC', 'CD', 'TN'],
  146. index_unique="-",
  147. join = 'outer'
  148. )
  150. # %%
  151. # Plot QC parameters before QC
  153. fig, axes = plt.subplots(nrows = 3, ncols = 1, figsize= (14,10))
  154. for i in range(3):
  155. ax = sc.pl.violin(concat_data, keys = qc_lst[i], jitter = 0.4, show = False, groupby = 'sample', xlabel = None, size = 1, ylabel = '', ax= axes[i])
  156. ax.spines['right'].set_visible(False)
  157. ax.spines['top'].set_visible(False)
  158. ax.tick_params(axis='both', which='major', labelsize= axis_tick_fs)
  159. ax.grid(False)
  160. ax.set_title(f'{qc_lst[i]}', fontsize = title_fs)
  161. plt.subplots_adjust(hspace = 0.4)
  162. plt.savefig('Figures/20231019_res/vlnplot_beforeqc_aftrdbremove_allqcmetrics_3snrna_20231104.pdf', bbox_inches = 'tight')
  164. # %%
  165. # Spot filtering
  167. max_gene = 8000
  168. min_gene = 200
  169. max_count = 60000
  170. max_mito = 20
  171. for adata in adata_lst:
  172. sc.pp.filter_cells(adata, max_counts=max_count)
  173. sc.pp.filter_cells(adata, min_genes= min_gene)
  174. sc.pp.filter_cells(adata, max_genes= max_gene)
  176. for i in range(3):
  177. adata = adata_lst[i]
  178. adata = adata[adata.obs['pct_counts_mt'] < max_mito, ~(adata.var_names == "Gm42418")]
  179. adata_lst[i] = adata
  181. for i in range(3):
  182. print(f"Number of spots/genes for {samp_lst[i]} sample: {len(adata_lst[i].obs)}/{len(adata_lst[i].var)}")
  185. # %%
  186. concat_data = sc.concat(
  187. adata_lst,
  188. label="sample",
  189. keys=['HC', 'CD', 'TN'],
  190. index_unique="-",
  191. join = 'outer'
  192. )
  194. concat_data
  195. # 24735 spots, 23362 genes
  196. # 24303 spots, 23267 genes
  198. # %%
  199. # Plot QC parameters before QC
  201. fig, axes = plt.subplots(nrows = 3, ncols = 1, figsize= (14,10))
  202. for i in range(3):
  203. ax = sc.pl.violin(concat_data, keys = qc_lst[i], jitter = 0.4, show = False, groupby = 'sample', xlabel = None, size = 1, ylabel = '', ax= axes[i])
  204. ax.spines['right'].set_visible(False)
  205. ax.spines['top'].set_visible(False)
  206. ax.tick_params(axis='both', which='major', labelsize= axis_tick_fs)
  207. ax.grid(False)
  208. ax.set_title(f'{qc_lst[i]}', fontsize = title_fs)
  209. plt.subplots_adjust(hspace = 0.4)
  210. plt.savefig('Figures/20231019_res/vlnplot_afterqc_aftrdbremove_allqcmetrics_3snrna_20231104.pdf', bbox_inches = 'tight')
  212. # %%
  213. # save raw count
  214. concat_data.write_h5ad('Tables/Data/snrna_dcn_3samples_int_afterdoubletrm_rawcount_outerjoin_20231104.h5ad')
  216. # %%
  217. #concat_data = sc.read_h5ad('Tables/Tables/h5ad_data/snrna_dcn_3samples_int_afterdoubletrm_rawcount_outerjoin_20230831.h5ad')
  219. # %%
  220. concat_data.layers["counts"] = concat_data.X.copy()
  222. # %% [markdown]
  223. # # Normalize data
  225. # %%
  226. sc.pp.normalize_total(concat_data)
  227. sc.pp.log1p(concat_data)
  229. # %%
  230. concat_data.raw = concat_data
  232. # %% [markdown]
  233. # # SCVI analysis
  235. # %%
  236. # Selecting highly variable genes
  237. sc.pp.highly_variable_genes(
  238. concat_data,
  239. flavor="seurat_v3",
  240. n_top_genes=2000,
  241. layer="counts",
  242. batch_key="sample",
  243. subset=True,
  244. )
  246. # %%
  247. TF_CPP_MIN_LOG_LEVEL=0
  249. # %%
  250. scvi.model.SCVI.setup_anndata(concat_data, layer="counts", batch_key="sample")
  252. # %%
  253. vae = scvi.model.SCVI(concat_data, n_layers=2, n_latent=30, gene_likelihood="nb")
  254. vae.train()
  256. # %%
  257. SCVI_LATENT_KEY = "X_scVI"
  258. concat_data.obsm[SCVI_LATENT_KEY] = vae.get_latent_representation()
  260. # %%
  261. # Clustering
  262. sc.pp.neighbors(concat_data, use_rep="X_scVI")
  264. # %%
  265. sc.tl.leiden(concat_data, resolution = 0.3, key_added = 'leiden_0.3')
  266. sc.tl.leiden(concat_data, resolution = 0.4, key_added = 'leiden_0.4')
  267. sc.tl.leiden(concat_data, resolution = 0.5, key_added = 'leiden_0.5')
  268. sc.tl.leiden(concat_data, resolution = 0.6, key_added = 'leiden_0.6')
  269. sc.tl.leiden(concat_data, resolution = 0.7, key_added = 'leiden_0.7')
  270. sc.tl.leiden(concat_data, resolution = 0.8, key_added = 'leiden_0.8')
  271. sc.tl.leiden(concat_data, resolution = 0.9, key_added = 'leiden_0.9')
  272. sc.tl.leiden(concat_data, resolution = 1, key_added = 'leiden_1.0')
  273. sc.tl.leiden(concat_data, resolution = 1.1, key_added = 'leiden_1.1')
  274. sc.tl.leiden(concat_data, resolution = 1.2, key_added = 'leiden_1.2')
  275. sc.tl.leiden(concat_data, resolution = 1.3, key_added = 'leiden_1.3')
  276. sc.tl.leiden(concat_data, resolution = 1.4, key_added = 'leiden_1.4')
  277. sc.tl.leiden(concat_data, resolution = 1.5, key_added = 'leiden_1.5')
  279. # %%
  280. sc.tl.umap(concat_data)
  282. # %% [markdown]
  283. # ## Plotting
  285. # %% [markdown]
  286. # ### Plot doublet prediction
  287. # This process is conducted after doublet detection, and before removing doublet.
  289. # %%
  290. concat_data.obs['doublet_prediction'] = concat_data.obs['doublet_prediction'].astype('category')
  292. # %%
  293. concat_data[concat_data.obs['doublet_prediction'] == False].obs.value_counts('sample')
  295. # %%
  296. # Check doublet
  297. mpl.rcParams['figure.figsize'] = (7,6)
  298. ax = sc.pl.umap(concat_data, color = ['doublet_prediction'], show = False,legend_fontsize = legend_fs, size = 10, sort_order = False)
  300. ax.title.set_fontsize(title_fs)
  301. ax.xaxis.label.set_fontsize(axis_label_fs)
  302. ax.yaxis.label.set_fontsize(axis_label_fs)
  304. plt.savefig('Figures/20231019_res/umap_snrna_3samples_sample_doubletprediction_outerjoin_20231104.pdf', bbox_inches = 'tight')
  306. # %% [markdown]
  307. # ### Plotting UMAP
  309. # %%
  310. mpl.rcParams['figure.figsize'] = (5,5)
  311. ax = sc.pl.umap(concat_data, color = ['leiden_0.4','leiden_0.5', 'leiden_0.6', 'leiden_0.7', 'leiden_0.8', 'leiden_0.9', 'leiden_1.0',], ncols = 3, show = False, legend_fontsize = legend_fs, size = 10, sort_order = False,
  312. legend_loc = 'on data')
  314. for sp in ax:
  315. sp.title.set_fontsize(title_fs)
  316. sp.xaxis.label.set_fontsize(axis_label_fs)
  317. sp.yaxis.label.set_fontsize(axis_label_fs)
  319. #plt.savefig('Figures/20230827_res/umap_snrna_3samples_rmdoublets_sample_leiden_multirestest_outerjoin_20230831.pdf', bbox_inches = 'tight')
  321. # %%
  322. concat_data.obs['leiden'] = concat_data.obs['leiden_1.0'].astype('category')
  324. # %%
  325. mpl.rcParams['figure.figsize'] = (5.2,5)
  326. ax = sc.pl.umap(concat_data, color = ['leiden'], show = False, legend_fontsize = legend_fs, size = 10, sort_order = False,
  327. wspace = 0.25, legend_loc = 'on data')
  329. ax.title.set_fontsize(title_fs)
  330. ax.xaxis.label.set_fontsize(axis_label_fs)
  331. ax.yaxis.label.set_fontsize(axis_label_fs)
  333. plt.savefig('Figures/20231019_res/umap_snrna_3samples_leiden1.0_20231104.pdf', bbox_inches = 'tight')
  335. # %%
  336. pd.crosstab(concat_data.obs['leiden'], concat_data.obs['sample'])
  338. # %%
  339. # violin plot of QC metrics for each leiden cluster
  340. fig, axes = plt.subplots(nrows = 3, ncols = 1, figsize= (14,10))
  341. for i in range(3):
  342. ax = sc.pl.violin(concat_data, keys = qc_lst[i], jitter = 0.4, show = False, groupby = 'leiden', xlabel = None, size = 1, ylabel = '', ax= axes[i])
  343. ax.spines['right'].set_visible(False)
  344. ax.spines['top'].set_visible(False)
  345. ax.tick_params(axis='both', which='major', labelsize= axis_tick_fs)
  346. ax.grid(False)
  347. ax.set_title(f'{qc_lst[i]}', fontsize = title_fs)
  348. plt.subplots_adjust(hspace = 0.4)
  349. plt.savefig('Figures/20231019_res/vlnplot_allqcmetrics_byleiden_res0.8_3snrna_20231104.pdf', bbox_inches = 'tight')
  351. # %%
  352. # plot UMAP based on QC metrics
  353. mpl.rcParams['figure.figsize'] = (5,5)
  354. ax = sc.pl.umap(concat_data, color = qc_lst, size = 15, show = False, legend_fontsize= legend_fs,legend_loc = 'on data')
  356. for sp in ax:
  357. sp.title.set_fontsize(title_fs)
  358. sp.xaxis.label.set_fontsize(axis_label_fs)
  359. sp.yaxis.label.set_fontsize(axis_label_fs)
  361. plt.savefig('Figures/20231019_res/umap_leiden_qccriteria_res1.0_20231104.pdf', bbox_inches= 'tight')
  363. # %%
  364. # Plot each samples individually
  365. fig, axs = plt.subplots(ncols = 3, nrows = 1, figsize = (12,3.5))
  367. sc.pl.umap(concat_data[concat_data.obs['sample'] == 'HC',], color="leiden_1.0", size = 10,legend_loc= None,
  368. title = 'HC', show = False, ax = axs[0])
  369. sc.pl.umap(concat_data[concat_data.obs['sample'] == 'CD',], color="leiden_1.0", size = 10,legend_loc= None,
  370. title = 'CD', show = False, ax = axs[1])
  371. sc.pl.umap(concat_data[concat_data.obs['sample'] == 'TN',], color="leiden_1.0", size = 10,legend_loc= None,
  372. title = 'TN', show = False, ax = axs[2])
  374. for subplot in axs:
  375. subplot.set_xlabel(subplot.get_xlabel(), fontsize= axis_label_fs)
  376. subplot.set_ylabel(subplot.get_ylabel(), fontsize= axis_label_fs)
  377. subplot.set_title(subplot.get_title(), fontsize = title_fs)
  379. plt.savefig("Figures/20231019_res/umap_3samples_snrna_rmdoublets_scviclusters_splitsample_outerjoin_20231104.pdf", bbox_inches = 'tight')
  381. # %%
  382. concat_data.obs['leiden'].value_counts()
  384. # %%
  385. # remove cluster 26 due to low nuclei count
  386. concat_data = concat_data[~(concat_data.obs['leiden'] == '26')].copy()
  388. # %%
  389. # export data
  390. concat_data.write_h5ad('Tables/Data/snrna_dcn_3samples_rmdoublet_intscvi_res1.0_20231104.h5ad')
  392. # %%
  393. #concat_data = sc.read_h5ad('Tables/Data/snrna_dcn_3samples_rmdoublet_intscvi_res1.0_20231104.h5ad')
  395. # %% [markdown]
  396. # # Find markers of each cluster (for annotation)
  398. # %%
  399. concat_data = concat_data.raw.to_adata()
  401. # %%
  402. concat_data
  403. # 24303 cells, 23267 genes
  405. # %%
  406. raw_adata = sc.read_h5ad('Tables/Data/snrna_dcn_3samples_int_afterdoubletrm_rawcount_outerjoin_20231104.h5ad')
  408. # %%
  409. concat_data.layers['counts'] = raw_adata.X.copy()
  411. # %%
  412. sc.tl.rank_genes_groups(concat_data, groupby = 'leiden', pts = True, method = 'wilcoxon')
  414. # %%
  415. markers_df = sc.get.rank_genes_groups_df(concat_data, group = None)
  417. # %%
  418. markers_df.to_csv('Tables/snrna_3samples_resolution1.0_markergenes_df_testwilcox_20231116.csv')
  420. # %% [markdown]
  421. # # Annotation using marker genes
  423. # %%
  424. concat_data.obs['cell_type'] = ''
  425. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['0','1','6','15'])] = 'Oligodendrocyte'
  426. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['2','4', '24'])] = 'Granule_cell'
  427. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['3','8'])] = 'Astrocyte'
  428. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['5','9', '14', '19', '20', '23'])] = 'Inh_DCN'
  429. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['7', '12'])] = 'Exc_DCN'
  430. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['10'])] = 'OPC'
  431. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['11'])] = 'Microglia'
  432. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['13','22'])] = 'Vascular_cell'
  433. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['16'])] = 'Ependymal_cell'
  434. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['17'])] = 'Bergmanns_glia'
  435. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['18'])] = 'Endothelial_cell'
  436. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['21'])] = 'UBC'
  437. #concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['24'])] = 'Granule_cell_TN'
  438. concat_data.obs['cell_type'][concat_data.obs['leiden'].isin(['25'])] = 'Purkinje'
  440. # %%
  441. concat_data.obs['cell_type'] = concat_data.obs['cell_type'].astype('category')
  443. # %%
  444. concat_data.obs['cell_type'].value_counts()
  446. # %%
  447. concat_data.write_h5ad('Tables/Data/snrna_dcn_3samples_rmdoublet_res1.0_afterannot_20231104.h5ad')
  449. # %%
  450. concat_data = sc.read_h5ad('Tables/Data/snrna_dcn_3samples_rmdoublet_res1.0_afterannot_20231104.h5ad')
  452. # %%
  453. concat_data.X = concat_data.layers['counts'].copy()
  454. concat_data.write_h5ad('Tables/Data/snrna_dcn_3samples_res1.0_afterannot_rawcnts_20240214.h5ad')
  456. # %%
  457. cnt_data = sc.read_h5ad('Tables/Data/snrna_dcn_3samples_res1.0_afterannot_rawcnts_20240214.h5ad')
  459. # %%
  460. cnt_data.obs.to_csv('Tables/Data/snrna_metadata.csv')
  462. # %%
  463. df = pd.DataFrame(cnt_data.X.toarray(), index=cnt_data.obs_names, columns=cnt_data.var_names)
  465. # Save the DataFrame to a CSV file
  466. df.to_csv('Tables/Data/snrna_onlycnt_mtx.csv')
  468. # %%
  469. mpl.rcParams['figure.figsize'] = (6.5,5.5)
  470. ax = sc.pl.umap(concat_data, color = ['leiden','cell_type'], size = 15, show = False, legend_fontsize= legend_fs,legend_loc = 'on data')
  472. for sp in ax:
  473. sp.title.set_fontsize(title_fs)
  474. sp.xaxis.label.set_fontsize(axis_label_fs)
  475. sp.yaxis.label.set_fontsize(axis_label_fs)
  477. plt.savefig('Figures/20231019_res/umap_leiden_celltype_colors_res1.0_20231104.pdf', bbox_inches= 'tight')
  479. # %%
  480. mpl.rcParams['figure.figsize'] = (6.5,5.5)
  481. ax = sc.pl.umap(concat_data, color = ['leiden','cell_type'], size = 15, show = False, legend_fontsize= legend_fs,legend_loc = 'on data')
  483. for sp in ax:
  484. sp.title.set_fontsize(title_fs)
  485. sp.xaxis.label.set_fontsize(axis_label_fs)
  486. sp.yaxis.label.set_fontsize(axis_label_fs)
  488. plt.savefig('Figures/20240604_res/umap_leiden_celltype_colors_res1.0_20240714.pdf', bbox_inches= 'tight')
  490. # %% [markdown]
  491. # ## Check marker genes of cell types
  493. # %%
  494. concat_data
  496. # %%
  497. hc_data = concat_data[concat_data.obs['sample'] == 'HC'].copy()
  498. hc_data
  500. # %%
  501. concat_data.uns['cell_type_colors']
  503. # %%
  504. hc_data.uns['cell_type_colors']
  506. # %%
  507. np.max(hc_data.X)
  509. # %%
  510. plt.rcParams['figure.figsize'] = (10,4.5)
  512. sc.pl.violin(hc_data, keys=['Eef1a2'], groupby='cell_type',
  513. #palette = ['#1f77b4', '#aa40fc', '#e377c2', '#b5bd61','#17becf', '#aec7e8'],
  514. show = False)
  516. plt.title('Eef1a2 expression in homecage (baseline)', size = title_fs)
  517. plt.gca().xaxis.label.set_fontsize(axis_label_fs)
  518. plt.gca().yaxis.label.set_fontsize(axis_label_fs)
  519. plt.tick_params(axis='both', which='major', labelsize= axis_tick_fs)
  520. plt.xticks(rotation = 15)
  522. plt.savefig('Figures/20240604_res/vlnplot_hc_eef1a2_expr_percelltype_all_20240715.pdf', bbox_inches = 'tight')
  524. # %%
  525. sc.tl.rank_genes_groups(hc_data, groupby = 'cell_type', pts = True, method = 'wilcoxon', key_added = 'celltype_markers')
  526. markers_df_hc = sc.get.rank_genes_groups_df(hc_data, group = None, key = 'celltype_markers')
  528. # %%
  529. markers_df.to_csv('Tables/snrna_3samples_celltype_markergenes_df_testwilcox_20231104.csv')
  531. # %%
  532. concat_data.obs['cell_type'] = pd.Categorical(concat_data.obs['cell_type'], categories = ['Astrocyte', 'Bergmanns_glia', 'Endothelial_cell', 'Ependymal_cell',
  533. 'Granule_cell', 'Granule_cell_TN', 'Exc_DCN', 'Inh_DCN','Microglia',
  534. 'OPC', 'Oligodendrocyte','Purkinje', 'UBC', 'Vascular_cell'], ordered = True)
  536. # %%
  537. sc.pl.stacked_violin(concat_data, groupby = 'cell_type', var_names = ['Aldoc', 'Gdf10', 'Cldn5', 'Foxj1', 'Gabra6', 'Slc17a6', 'Gad2', 'Tmem119', 'Pdgfra','Plp1', 'Pcp2',
  538. 'Eomes', 'Vtn'], swap_axes = True, show = False,figsize = (10,10),
  539. save = 'celltype_markers_20231104.pdf'
  540. )
  542. # %%
  543. markers_df_hc[(markers_df_hc['names'] == 'Eef1a2')].sort_values('logfoldchanges', ascending = False)
  545. # %%
  546. pd.crosstab(concat_data.obs['sample'], concat_data.obs['cell_type'])
  548. # %% [markdown]
  549. # # Subset and reclustering of DCN inhibitory neurons
  551. # %%
  552. concat_data = sc.read_h5ad('Tables/Data/snrna_dcn_3samples_rmdoublet_res1.0_afterannot_20231104.h5ad')
  554. # %%
  555. adata_inh = concat_data[concat_data.obs['cell_type'].isin(['Inh_DCN'])].copy()
  557. # %%
  558. adata_inh.X = adata_inh.layers['counts'].copy()
  560. # %%
  561. adata_inh
  562. # 2032 cells, 23267 genes
  564. # %% [markdown]
  565. # ## QC
  567. # %%
  568. inh_lst = []
  570. for samp in samp_lst:
  571. adata = adata_inh[adata_inh.obs['sample'] == samp].copy()
  572. inh_lst.append(adata)
  574. # %%
  575. for adata in inh_lst:
  576. sc.pp.filter_genes(adata, min_cells = 3)
  578. # %%
  579. inh_lst
  581. # %%
  582. adata_inh = sc.concat(
  583. inh_lst,
  584. join = 'outer'
  585. )
  587. # %%
  588. adata_inh
  589. # 2032 nuclei, 18842 genes
  591. # %% [markdown]
  592. # ## Normalize
  594. # %%
  595. sc.pp.normalize_total(adata_inh)
  596. sc.pp.log1p(adata_inh)
  597. adata_inh.raw = adata_inh
  599. # %% [markdown]
  600. # ## Run SCVI
  602. # %%
  603. # Selecting highly variable genes
  604. sc.pp.highly_variable_genes(
  605. adata_inh,
  606. flavor="seurat_v3",
  607. n_top_genes=2000,
  608. layer="counts",
  609. batch_key="sample",
  610. subset=True,
  611. )
  612. TF_CPP_MIN_LOG_LEVEL=0
  613. scvi.model.SCVI.setup_anndata(adata_inh, layer="counts", batch_key="sample")
  614. vae_dcn = scvi.model.SCVI(adata_inh, n_layers=2, n_latent=30, gene_likelihood="nb")
  615. vae_dcn.train()
  617. # %%
  618. SCVI_LATENT_KEY = "X_scVI"
  619. adata_inh.obsm[SCVI_LATENT_KEY] = vae_dcn.get_latent_representation()
  621. # %%
  622. # Clustering
  623. sc.pp.neighbors(adata_inh, use_rep="X_scVI", n_pcs = 20)
  625. # %%
  626. sc.tl.umap(adata_inh)
  628. # %%
  629. sc.tl.leiden(adata_inh, resolution = 0.3, key_added = 'leiden_0.3')
  630. sc.tl.leiden(adata_inh, resolution = 0.4, key_added = 'leiden_0.4')
  631. sc.tl.leiden(adata_inh, resolution = 0.5, key_added = 'leiden_0.5')
  632. sc.tl.leiden(adata_inh, resolution = 0.6, key_added = 'leiden_0.6')
  633. sc.tl.leiden(adata_inh, resolution = 0.7, key_added = 'leiden_0.7')
  634. sc.tl.leiden(adata_inh, resolution = 0.8, key_added = 'leiden_0.8')
  635. sc.tl.leiden(adata_inh, resolution = 0.9, key_added = 'leiden_0.9')
  636. sc.tl.leiden(adata_inh, resolution = 1, key_added = 'leiden_1.0')
  637. sc.tl.leiden(adata_inh, resolution = 1.1, key_added = 'leiden_1.1')
  638. sc.tl.leiden(adata_inh, resolution = 1.2, key_added = 'leiden_1.2')
  639. sc.tl.leiden(adata_inh, resolution = 1.3, key_added = 'leiden_1.3')
  640. sc.tl.leiden(adata_inh, resolution = 1.4, key_added = 'leiden_1.4')
  641. sc.tl.leiden(adata_inh, resolution = 1.5, key_added = 'leiden_1.5')
  642. sc.tl.leiden(adata_inh, resolution = 1.6, key_added = 'leiden_1.6')
  643. sc.tl.leiden(adata_inh, resolution = 1.7, key_added = 'leiden_1.7')
  644. sc.tl.leiden(adata_inh, resolution = 1.8, key_added = 'leiden_1.8')
  645. sc.tl.leiden(adata_inh, resolution = 1.9, key_added = 'leiden_1.9')
  646. sc.tl.leiden(adata_inh, resolution = 2.0, key_added = 'leiden_2.0')
  648. # %%
  649. sc.pl.umap(adata_inh, color = 'leiden_1.0')
  651. # %%
  652. plt.rcParams['figure.figsize'] = (12,8)
  653. sc.pl.umap(adata_inh, color=["leiden_0.3", "leiden_0.5", "leiden_0.6", "leiden_0.7", 'leiden_1.4'], size = 60,ncols = 3, legend_fontsize = legend_fs, show = False,
  654. wspace = 0.2)
  656. # %%
  657. sc.pl.dendrogram(adata_inh, groupby = 'leiden_0.3')
  659. # %%
  660. adata_inh.write_h5ad('Tables/Data/snrna_subsetdcn_afterqc_scvi_20231104.h5ad')
  662. # %%
  663. #adata_inh = sc.read_h5ad('Tables/Data/snrna_subsetdcn_afterqc_scvi_20231104.h5ad')
  665. # %%
  666. adata_inh.X = adata_inh.layers['counts'].copy()
  668. # %% [markdown]
  669. # ## Annotating clusters using SCANVI
  671. # %%
  672. adata_inh
  674. # %%
  675. adata_inh = adata_inh.raw.to_adata()
  677. # %%
  678. adata_raw = sc.read_h5ad('Tables/Data/snrna_dcn_3samples_int_afterdoubletrm_rawcount_outerjoin_20231104.h5ad')
  679. adata_raw_inh = adata_raw[adata_inh.obs_names,adata_inh.var_names]
  681. # %%
  682. adata_inh.layers['counts'] = adata_raw_inh.X.copy()
  683. adata_inh.layers['log1p'] = adata_inh.X.copy()
  684. adata_inh.X = adata_inh.layers['counts'].copy()
  686. # %% [markdown]
  687. # Reference: Kebschull et al., 2021
  689. # %%
  690. # Train reference
  691. adata_ref = sc.read_h5ad('/data1/Spatial_DCN/Resources/kebschull_dcn_rawcounts_anndata.h5ad')
  692. adata_ref
  693. # 4605 cells, 53797 genes
  695. # %%
  696. adata_ref = adata_ref[adata_ref.obs['final.clusters2'].str.contains('Inh')].copy()
  697. adata_ref
  698. # 2363 cells, 53797 genes
  700. # %%
  701. adata_ref.layers['counts'] = adata_ref.X.copy()
  703. # %%
  704. # Match sample column name to reference
  705. adata_inh.obs['orig.ident'] = adata_inh.obs['sample']
  707. # %%
  708. # concatenate dataset with reference
  709. merged_data = adata_inh.concatenate(adata_ref)
  711. # %%
  712. merged_data
  713. # 4395 cells, 18020 genes
  715. # %% [markdown]
  716. # ### Run SCVI on reference merged dataset
  718. # %%
  719. merged_data.layers['counts'] = merged_data.X.copy()
  721. sc.pp.normalize_total(merged_data, target_sum = 1e4)
  722. sc.pp.log1p(merged_data)
  723. merged_data.raw = merged_data
  725. sc.pp.highly_variable_genes(
  726. merged_data, n_top_genes=3000, batch_key="orig.ident", subset=True, layer = 'counts', flavor = 'seurat_v3'
  727. )
  729. # %%
  730. TF_CPP_MIN_LOG_LEVEL=0
  732. # %%
  733. scvi.model.SCVI.setup_anndata(merged_data, batch_key="orig.ident", layer="counts", categorical_covariate_keys= ['batch'])
  735. # %%
  736. vae = scvi.model.SCVI(merged_data)
  737. vae.train()
  739. # %%
  740. # set reference column to project
  741. merged_data.obs['final.clusters2'] = merged_data.obs['final.clusters2'].cat.add_categories('Unknown')
  742. merged_data.obs = merged_data.obs.fillna(value = {'final.clusters2': 'Unknown'})
  744. # %% [markdown]
  745. # ### train SCANVI
  747. # %%
  748. lvae = scvi.model.SCANVI.from_scvi_model(vae, adata = merged_data, unlabeled_category = 'Unknown',labels_key = 'final.clusters2')
  749. lvae.train(max_epochs = 20, n_samples_per_label = 100)
  751. # %%
  752. merged_data.obs['SCANVI_prediction'] = lvae.predict(merged_data)
  754. # %%
  755. merged_data.obs['bc2'] = merged_data.obs.index.map(lambda x: x[:-2])
  756. cell_mapper = dict(zip(merged_data.obs.bc2, merged_data.obs.SCANVI_prediction))
  758. adata_inh.obs['SCANVI_predicted'] = adata_inh.obs.index.map(cell_mapper)
  760. # %%
  761. sc.pl.dendrogram(adata_inh, groupby = 'leiden_0.3')
  763. # %%
  764. # Clustering
  765. #mpl.rcParams['figure.figsize'] = (5.5,5)
  766. sc.pl.umap(adata_inh, color = ['leiden_0.5','SCANVI_predicted'], wspace = 0.35)
  768. # %% [markdown]
  769. # ## Find markers of DCN neuronal celltypes
  771. # %%
  772. sc.tl.rank_genes_groups(adata_inh, groupby = 'leiden_0.3', pts = True, method = 'wilcoxon', key_added='leiden_0.3_markers')
  773. markers_df = sc.get.rank_genes_groups_df(adata_inh, group = None, key='leiden_0.3_markers')
  775. # %%
  776. sc.tl.rank_genes_groups(adata_inh, groupby = 'neuronal_celltype', pts = True, method = 'wilcoxon', key_added='celltype_markers')
  777. markers_df = sc.get.rank_genes_groups_df(adata_inh, group = None, key='celltype_markers')
  779. # %%
  780. markers_df[markers_df['names'] == 'Kit'].sort_values(by = 'logfoldchanges', ascending = False)
  782. # %%
  783. markers_df[markers_df['names'] == 'Sox14'].sort_values(by = 'logfoldchanges', ascending = False)
  785. # %%
  786. marker_df = markers_df.rename(columns={"group": "Neuronal_celltype", "names": "Gene"})
  788. # %%
  789. adata_inh.obs['leiden_0.3'].value_counts()
  791. # %%
  792. adata_inh = adata_inh[~(adata_inh.obs['leiden_0.3'] == '7')]
  794. # %%
  795. # annotate inhibitory neuron cell type
  796. adata_inh.obs['neuronal_celltype'] = ''
  797. adata_inh.obs['neuronal_celltype'][adata_inh.obs['leiden_0.3'].isin(['1','2','3'])] = 'Inh_Zfhx4'
  798. adata_inh.obs['neuronal_celltype'][adata_inh.obs['leiden_0.3'].isin(['0', '4', '6', '8'])] = 'Inh_Kit'
  799. adata_inh.obs['neuronal_celltype'][adata_inh.obs['leiden_0.3'].isin(['5'])] = 'Inh_Piezo2'
  801. adata_inh.obs['neuronal_celltype'] = adata_inh.obs['neuronal_celltype'].astype('category')
  803. # %%
  804. adata_inh.obs['leiden'] = adata_inh.obs['leiden_0.3'].astype('category')
  806. # %%
  807. mpl.rcParams['figure.figsize'] = (5,5)
  808. ax = sc.pl.umap(adata_inh, color = ['leiden', 'neuronal_celltype'], wspace = 0.24, show = False,
  809. legend_fontsize= legend_fs, size = 40)
  811. for sp in ax:
  812. sp.title.set_fontsize(title_fs)
  813. sp.xaxis.label.set_fontsize(axis_label_fs)
  814. sp.yaxis.label.set_fontsize(axis_label_fs)
  816. plt.savefig('Figures/20240217_res/umap_snrna_onlyinh_color_leiden_ct_20240311.pdf', bbox_inches = 'tight')
  818. # %%
  819. figs, axs = plt.subplots(ncols = 1, nrows = 3, figsize = (4.2, 7))
  821. sc.pl.violin(adata_inh, groupby='neuronal_celltype', keys=['Kit'], ax = axs[0], show = False, size = 0)
  822. axs[0].set_title('Kit', size = title_fs, x = 0.08)
  824. sc.pl.violin(adata_inh, groupby='neuronal_celltype', keys=['Piezo2'], ax = axs[1], show = False, size = 0)
  825. axs[1].set_title('Piezo2', size = title_fs, x = 0.1)
  827. sc.pl.violin(adata_inh, groupby='neuronal_celltype', keys=['Zfhx4'], ax = axs[2], show = False, size = 0)
  828. axs[2].set_title('Zfhx4', size = title_fs, x = 0.1)
  830. for ax in axs:
  831. ax.title.set_fontsize(title_fs)
  832. ax.set_xlabel('')
  833. ax.set_ylabel('')
  834. ax.tick_params(labelsize=axis_tick_fs)
  836. plt.subplots_adjust(hspace = 0.4)
  837. plt.savefig('Figures/20240217_res/vlnplot_inhibitory_celltype_markers_20240507.pdf', bbox_inches = 'tight')
  839. # %%
  840. # Plot each samples individually
  841. fig, axs = plt.subplots(ncols = 3, nrows = 1, figsize = (11.5,3))
  843. sc.pl.umap(adata_inh[adata_inh.obs['sample'] == 'HC',], color="leiden", size = 35,legend_loc= None,
  844. title = 'HC', show = False, ax = axs[0])
  845. sc.pl.umap(adata_inh[adata_inh.obs['sample'] == 'CD',], color="leiden", size = 35,legend_loc= None,
  846. title = 'CD', show = False, ax = axs[1])
  847. sc.pl.umap(adata_inh[adata_inh.obs['sample'] == 'TN',], color="leiden", size = 35,legend_loc= None,
  848. title = 'TN', show = False, ax = axs[2])
  850. for subplot in axs:
  851. subplot.set_xlabel(subplot.get_xlabel(), fontsize= axis_label_fs)
  852. subplot.set_ylabel(subplot.get_ylabel(), fontsize= axis_label_fs)
  853. subplot.set_title(subplot.get_title(), fontsize = title_fs)
  855. plt.savefig('Figures/20240217_res/umap_snrna_onlyinh_splitsample_20240311.pdf', bbox_inches = 'tight')
  857. # %%
  858. adata_inh.X = adata_inh.layers['log1p'].copy()
  860. # %%
  861. # Plot each samples individually
  862. fig, axs = plt.subplots(ncols = 3, nrows = 1, figsize = (11.5,3))
  864. sc.pl.umap(adata_inh[adata_inh.obs['sample'] == 'HC',], color="Grm5", size = 35,legend_loc= None,
  865. title = 'HC', show = False, ax = axs[0], vmax = 3.5)
  866. sc.pl.umap(adata_inh[adata_inh.obs['sample'] == 'CD',], color="Grm5", size = 35,legend_loc= None,
  867. title = 'CD', show = False, ax = axs[1])
  868. sc.pl.umap(adata_inh[adata_inh.obs['sample'] == 'TN',], color="Grm5", size = 35,legend_loc= None,
  869. title = 'TN', show = False, ax = axs[2], vmax = 3.5)
  871. for subplot in axs:
  872. subplot.set_xlabel(subplot.get_xlabel(), fontsize= axis_label_fs)
  873. subplot.set_ylabel(subplot.get_ylabel(), fontsize= axis_label_fs)
  874. subplot.set_title(subplot.get_title(), fontsize = title_fs)
  876. #plt.savefig('Figures/20240217_res/umap_snrna_onlyinh_splitsample_20240311.pdf', bbox_inches = 'tight')
  878. # %%
  879. adata_inh.write_h5ad('Tables/Data/snrna_subsetinh_afterannot_20240311.h5ad')
  881. # %%
  882. adata_inh = sc.read_h5ad('Tables/Data/snrna_subsetinh_afterannot_20240311.h5ad')
  884. # %%
  885. markers_df.to_csv('Tables/3samples_dcn_inh_neurocelltype_markers_20241022.csv')
  887. # %% [markdown]
  888. # # Export for MAST
  890. # %%
  891. concat_out = concat_data[~(concat_data.obs['cell_type'].isin(['Granule_cell', 'UBC','Purkinje', 'Bergmanns_glia','Ependymal_cell', 'Vascular_cell', 'Endothelial_cell']))].copy()
  892. #dcn_out = adata_inh[~(adata_inh.obs['neuronal_celltype'] == 'Inhibitory_1_Zfhx4_TN')].copy()
  895. # %%
  896. concat_out.obs['cell_type'].unique()
  898. # %%
  899. concat_out.X = concat_out.layers['counts'].copy()
  900. #dcn_out.X = dcn_out.layers['counts'].copy()
  902. # %%
  903. sc.pp.normalize_total(concat_out, target_sum=1e6)
  904. sc.pp.log1p(concat_out)
  906. #
  907. #
  908. # ize_total(dcn_out, target_sum=1e6)
  909. #sc.pp.log1p(dcn_out)
  911. # %%
  912. def prep_anndata(adata_):
  913. def fix_dtypes(adata_):
  914. df = pd.DataFrame(adata_.X.A, index=adata_.obs_names, columns=adata_.var_names)
  915. df = df.join(adata_.obs)
  916. return sc.AnnData(df[adata_.var_names], obs=df.drop(columns=adata_.var_names))
  918. adata_ = fix_dtypes(adata_)
  919. sc.pp.filter_genes(adata_, min_cells=3)
  920. return adata_
  922. # %%
  923. concat_out = prep_anndata(concat_out)
  924. #dcn_out = prep_anndata(dcn_out)
  926. # %%
  927. concat_out
  928. #17383 cells, 22728 genes
  930. # %%
  931. adata_lst = []
  932. for ct in concat_out.obs['cell_type'].unique():
  933. adata_ct = concat_out[concat_out.obs['cell_type'] == ct].copy()
  934. adata_ct = prep_anndata(adata_ct)
  935. print(f"{ct}: {len(adata_ct.obs)} nucleis, {len(adata_ct.var)} genes")
  936. adata_lst.append(adata_ct)
  938. # %%
  939. inh_out = adata_inh.copy()
  941. # %%
  942. inh_out.X = adata_inh.layers['counts'].copy()
  944. # %%
  945. def prep_anndata(adata_):
  946. def fix_dtypes(adata_):
  947. df = pd.DataFrame(adata_.X.A, index=adata_.obs_names, columns=adata_.var_names)
  948. df = df.join(adata_.obs)
  949. return sc.AnnData(df[adata_.var_names], obs=df.drop(columns=adata_.var_names))
  951. adata_ = fix_dtypes(adata_)
  952. sc.pp.filter_genes(adata_, min_cells=3)
  953. return adata_
  955. # %%
  956. adata_lst = []
  957. for ct in inh_out.obs['neuronal_celltype'].unique():
  958. adata_ct = inh_out[inh_out.obs['neuronal_celltype'] == ct].copy()
  959. sc.pp.normalize_total(adata_ct, target_sum=1e6)
  960. sc.pp.log1p(adata_ct)
  961. adata_ct = prep_anndata(adata_ct)
  962. print(f"{ct}: {len(adata_ct.obs)} nucleis, {len(adata_ct.var)} genes")
  963. adata_lst.append(adata_ct)
  965. # %%
  966. for ad in adata_lst:
  967. ad.write_h5ad(f"Tables/Data/MAST_adatas/adata_{ad.obs['neuronal_celltype'][0]}_forMAST_analysis_20240311.h5ad")
  969. # %%
  970. np.max(adata_lst[0].X)

preprocess_annotate_snrna.ipynb, under CC-BY-4.0 · at the source

Overview

Authors: Jungeun Ji1,2, Jinhee Baek3,4, Kyoung-Doo Hwang3,4, Seunghwan Choi5, Junko Kasuya6, Seung-Eon Roh7, Sang Jeong Kim3,4,8,9, Ted Abel6, Joon-Yong An1,2,5, Yong-Seok Lee3,4,8,9,10
ORCID iDs: Ted Abel, Joon-Yong An, Yong-Seok Lee
  1. Department of Integrated Biomedical and Life Science, Korea University,Seoul, Republic of Korea
  2. BK21FOUR R&E Center for Learning Health Systems, Korea University,Seoul, Republic of Korea
  3. Department of Physiology, Seoul National University College of Medicine,Seoul, Republic of Korea
  4. Department of Biomedical Science, Seoul National University College of Medicine,Seoul, Republic of Korea
  5. School of Biosystems and Biomedical Sciences, College of Health Sciences, Korea University,Seoul, Republic of Korea
  6. Department of Neuroscience and Pharmacology, Iowa Neuroscience Institute, University of Iowa,Iowa City, IA USA
  7. Department of Neuroscience, Johns Hopkins University,Baltimore, MD USA
  8. Neuroscience Research Institute, Seoul National Medical Research Center, Seoul, Republic of Korea
  9. Wide River Institute of Immunology, Seoul National University,Hongcheon, Republic of Korea
  10. Convergence Dementia Research Center, Medical Research Center, Seoul National University,Seoul, Republic of Korea
Institutions: Korea University (South Korea); Seoul National University (South Korea); University of Iowa (United States); Johns Hopkins University (United States)
Journal: Communications biology, volume 9, issue 1, article 878
Dates: received 17 April 2025; accepted 30 March 2026; published online 24 April 2026
Type: Research article · Language: English
License: CC BY-NC-ND
Identifiers: DOI 10.1038/s42003-026-10034-0 · PMID 42032255 · PMCID PMC13319209 · OpenAlex W7155565318
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: genetics / omics (modality), mouse (organism), cellular / molecular (subfield)
Methods: Statistics, fMRI & imaging
Keywords: Molecular neuroscience, Learning and memory
MeSH: Cerebellum*, Conditioning, Classical*, Fear*, Gene Expression Regulation*, Animals, Male, Memory, Mice, Mice, Inbred C57BL, Neurons (* major topic)
Topic: Memory and Neural Mechanisms (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: Korea Health Industry Development Institute (RS-2025-25459221); National Research Foundation of Korea (RS-2024-00439474)
Citations: not cited yet (Europe PMC); 86 references in the paper

Abstract

The abstract is not reproduced here: the paper's license (CC BY-NC-ND) does not allow it. Read it in the paper, at the publisher or on Europe PMC.

Repository

Its files are read in the Code ↔ Paper reader above, with 17 matches between paragraphs and lines of code.

Zenodo 15335138

License: CC-BY-4.0
State: the link answers, verified on 30 September 2026
Evidence: files inventoried
Size: 1 file
Software Heritage: not checked
Found in: “Code availability”
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: Seurat (8 files), tidyverse (8 files), cowplot (7 files), Matplotlib (5 files), NumPy (5 files), pandas (5 files), Scanpy (5 files), seaborn (5 files), PyTorch (4 files), ggplot2 (3 files), anndata (2 files), clusterProfiler (2 files), ComplexHeatmap (2 files), limma (2 files), SingleCellExperiment (2 files), circlize (1 file), data.table (1 file), DESeq2 (1 file), ggpubr (1 file), reshape2 (1 file), reticulate (1 file), rstatix (1 file), SciPy (1 file), Squidpy (1 file), statsmodels (1 file)
Availability: 1 check, the latest on 30 September 2026: the link answers (HTTP 200)
  • 30 September 2026: the link answers (HTTP 200)
13 files
At the source: zenodo.org/records/15335138

Code availability statement

The paper has a code availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

Read it in the paper: doi.org/10.1038/s42003-026-10034-0.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 13 scripts, each with its path and the digest of its content;
  • 17 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Code and data availability statement

The paper has a code and data availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

Read it in the paper: doi.org/10.1038/s42003-026-10034-0.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 30 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 10 authors, 2 keywords, 10 MeSH terms, 2 funders, 86 references.

Cite

This paper

Ji, J., Baek, J., Hwang, K.-D., Choi, S., Kasuya, J., Roh, S.-E., Kim, S. J., Abel, T., An, J.-Y., & Lee, Y.-S. (2026). Region- and cell type-specific changes in gene expression in the cerebellum after classical fear conditioning. Communications biology, 9(1), 878. https://doi.org/10.1038/s42003-026-10034-0

BibTeX

@article{ji2026region,
author = {Ji, Jungeun and Baek, Jinhee and Hwang, Kyoung-Doo and Choi, Seunghwan and Kasuya, Junko and Roh, Seung-Eon and Kim, Sang Jeong and Abel, Ted and An, Joon-Yong and Lee, Yong-Seok},
title = {{Region- and cell type-specific changes in gene expression in the cerebellum after classical fear conditioning}},
journal = {Communications biology},
year = {2026},
month = apr,
volume = {9},
number = {1},
pages = {878},
publisher = {Nature Publishing Group},
issn = {2399-3642},
doi = {10.1038/s42003-026-10034-0},
url = {https://doi.org/10.1038/s42003-026-10034-0},
pmid = {42032255},
pmcid = {PMC13319209}
}

RIS

TY - JOUR
AU - Ji, Jungeun
AU - Baek, Jinhee
AU - Hwang, Kyoung-Doo
AU - Choi, Seunghwan
AU - Kasuya, Junko
AU - Roh, Seung-Eon
AU - Kim, Sang Jeong
AU - Abel, Ted
AU - An, Joon-Yong
AU - Lee, Yong-Seok
TI - Region- and cell type-specific changes in gene expression in the cerebellum after classical fear conditioning
T2 - Communications biology
J2 - Commun Biol
PY - 2026
DA - 2026/04/24
VL - 9
IS - 1
SP - 878
SN - 2399-3642
PB - Nature Publishing Group
DO - 10.1038/s42003-026-10034-0
UR - https://doi.org/10.1038/s42003-026-10034-0
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s42003-026-10034-0",
"type": "article-journal",
"title": "Region- and cell type-specific changes in gene expression in the cerebellum after classical fear conditioning",
"container-title": "Communications biology",
"author": [
{
"family": "Ji",
"given": "Jungeun"
},
{
"family": "Baek",
"given": "Jinhee"
},
{
"family": "Hwang",
"given": "Kyoung-Doo"
},
{
"family": "Choi",
"given": "Seunghwan"
},
{
"family": "Kasuya",
"given": "Junko"
},
{
"family": "Roh",
"given": "Seung-Eon"
},
{
"family": "Kim",
"given": "Sang Jeong"
},
{
"family": "Abel",
"given": "Ted"
},
{
"family": "An",
"given": "Joon-Yong"
},
{
"family": "Lee",
"given": "Yong-Seok"
}
],
"container-title-short": "Commun Biol",
"volume": "9",
"issue": "1",
"page": "878",
"DOI": "10.1038/s42003-026-10034-0",
"PMID": "42032255",
"PMCID": "PMC13319209",
"ISSN": "2399-3642",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s42003-026-10034-0",
"language": "en",
"issued": {
"date-parts": [
[
2026,
4,
24
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1016/j.xcrm.2026.102766 [code]
A longitudinal single-cell and spatial multiomic atlas of pediatric high-grade glioma.
Journal: Cell reports. Medicine
In common: SingleCellExperiment, limma, anndata, 17 other tools, genetics / omics, cellular / molecular, 5 references
[2] doi:10.1038/s41586-026-10629-x [code]
Whole-genome duplication shaped cell-type evolution in the vertebrate brain.
Journal: Nature
In common: reticulate, rstatix, anndata, 16 other tools, genetics / omics, mouse, cellular / molecular, 4 references
[3] doi:10.1186/s13059-026-04177-w [code]
Genomic sequence evolution underlying human neocortical interareal diversification.
Journal: Genome biology
In common: Squidpy, reticulate, rstatix, 15 other tools, genetics / omics, mouse, cellular / molecular, 4 references
[4] doi:10.1038/s41586-026-10214-2 [code]
Multidimensional profiling of heterogeneity in supratentorial ependymomas.
Journal: Nature
In common: Squidpy, SingleCellExperiment, reticulate, 18 other tools, genetics / omics, mouse
[5] doi:10.1002/imt2.70163 [code]
Spatial multi-omics unveils sphingolipid metabolic reprogramming within the retinal pathological niche.
Journal: iMeta
In common: SingleCellExperiment, limma, rstatix, 18 other tools, genetics / omics, mouse, cellular / molecular
[6] doi:10.1016/j.celrep.2026.117073 [code]
Single-cell epigenomics uncovers heterochromatin instability and transcription factor dysfunction during mouse brain aging.
Journal: Cell reports
In common: reticulate, anndata, DESeq2, 14 other tools, genetics / omics, mouse, cellular / molecular, 6 references
[7] doi:10.1038/s41467-026-71803-3 [code]
Charting the transition from in vitro gliogenesis to the in vivo maturation of human glial progenitor cells transplanted into the hypomyelinated mouse brain.
Journal: Nature communications
In common: Squidpy, reticulate, anndata, 14 other tools, genetics / omics, mouse, cellular / molecular, 4 references
[8] doi:10.1016/j.cpblue.2026.100007 [code]
An integrated single-cell and spatial proteotranscriptomics atlas of fibroblast-driven immunoregulation within the human adult oral cavity.
Journal: Cell press blue
In common: SingleCellExperiment, reticulate, limma, 17 other tools, 1 reference
[9] doi:10.1371/journal.pcbi.1014327 [code]
Supervised deep learning with gene functional annotation for cell classification.
Journal: PLoS computational biology
In common: SingleCellExperiment, reticulate, limma, 16 other tools, genetics / omics, cellular / molecular, 1 reference
[10] doi:10.1038/s41467-026-71525-6 [code]
Single-nucleus brain transcriptomics reveals microglia dysfunction in multiple system atrophy.
Journal: Nature communications
In common: rstatix, anndata, circlize, 15 other tools, genetics / omics, cellular / molecular, 2 references

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.