OSCR

Optimizing NGN2 Dosage Enhances the Neuronal Enrichment of iPSC-Derived Neuronal Cultures.

Code ↔ Paper

7 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 7 matches
  1. [1] § Experimental Procedures › Bioinformatic and Computational Analysis ↔ python/PCA and Volcano plot.ipynb, lines 1063–1097 · score 0.79 · gProfiler, CC, KEGG, MF, REAC, WP
  2. [2] § Experimental Procedures › Bioinformatic and Computational Analysis ↔ python/2025-ipscs-scp-github.ipynb, lines 220–223 · score 0.61 · UMAP, Scanpy, pp, resolution, neighbors, Leiden
  3. [3] § Results › Proteomic Analysis and Molecular Characterization of NGN2-Induced iPSC Transgenic Cultures ↔ python/Cluster heatmap and centroid distances.ipynb, lines 298–330 · score 0.60 · highlights small distances, pairwise Euclidean distances, clustering heatmap, proteomic
  4. [4] § Results › Proteomic Analysis and Molecular Characterization of NGN2-Induced iPSC Transgenic Cultures ↔ python/PCA and Volcano plot.ipynb, lines 718–744 · score 0.59 · ALDH1L1, CD44, MKI67, TUBB3, SOX2, SYN1
  5. [5] § Results › Proteomic Analysis and Molecular Characterization of NGN2-Induced iPSC Transgenic Cultures ↔ python/PCA and Volcano plot.ipynb, lines 1177–1261 · score 0.56 · GO term, gliogenesis, glutamatergic, astrocytes, volcano, enrichment
  6. [6] § Results › Proteomic Analysis and Molecular Characterization of NGN2-Induced iPSC Transgenic Cultures ↔ python/PCA and Volcano plot.ipynb, lines 1389–1457 · score 0.55 · Log2 fold change, Mann Whitney, NOTCH1, Volcano, bar, astrocytes
  7. [7] § Experimental Procedures › Linear Modeling ↔ python/Baseline linear model .ipynb, lines 120–156 · score 0.52 · linear model, OLS, ANOVA, fitted, row, Log2

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 · 1,620 lines · 49 KB · no license · 4 matches

  1. # %%
  2. import pandas as pd
  3. import plotly.express as px
  4. from sklearn.decomposition import PCA
  5. from sklearn.impute import KNNImputer
  6. import numpy as np
  7. import matplotlib.pyplot as plt
  8. import pandas as pd
  9. import seaborn as sns
  10. import matplotlib.pyplot as plt
  11. from scipy.spatial.distance import pdist, squareform
  13. # %%
  14. import os
  15. os.getcwd()
  17. # %%
  18. # Load your spreadsheet data into a Pandas DataFrame
  19. df = pd.read_csv(r"..\\data\2025\dox-mzml.pg_matrix (1).tsv", sep='\t', index_col=2)
  20. df
  21. df.drop(['Protein.Group', 'Protein.Names', 'First.Protein.Description','N.Sequences','N.Proteotypic.Sequences'],axis=1, inplace=True)
  22. df
  24. # %%
  25. df.columns = [col.split("\\")[-1].split(".")[0] for col in df.columns]
  26. df
  28. # %%
  29. np.log(df).boxplot(fontsize=5)
  30. #plt.axhline(df.count(axis=0), color='red', linestyle='--', label='Mean')
  31. plt.title('raw log2 quantities per sample')
  32. #plt.savefig(r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment/boxquant_raw.svg", bbox_inches='tight')
  33. plt.show()
  35. # %%
  36. df.count(axis=0)
  38. # %%
  39. plt.rcParams['figure.figsize'] = 3,3
  40. #plt.rcParams['axes.grid'] = False
  41. df.count(axis=0).hist(bins=20, grid=False)
  42. plt.axvline(x=7500, color='black', linewidth=3)
  43. plt.xlabel('number of protein IDs')
  44. plt.ylabel('count')
  45. #plt.savefig(r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment/proteins_per_condition.svg", bbox_inches='tight')
  46. plt.show()
  48. # %%
  49. df.T.index
  51. # %%
  52. dropthese = df.T.index[df.count()<7500]
  53. dropthese
  55. # %%
  56. df = df.drop(dropthese, axis=1)
  58. # %%
  59. df.count()
  61. # %%
  62. # Assuming your DataFrame is named 'df'
  63. column_sums = df.sum(axis=0) # Calculate the sum of each column
  65. # Divide each element by the corresponding column sum
  66. df_normalized = df.div(column_sums, axis=1)
  67. df_normalized
  69. # %%
  70. # Handle missing values (e.g., impute with mean or median), imputation may contain very small numbers
  71. # Create a KNN imputer object
  72. imputer = KNNImputer(n_neighbors=5) # Adjust n_neighbors as needed
  74. # Impute missing values using KNN
  75. df_imputed = imputer.fit_transform(df_normalized)
  77. # Create a new DataFrame with imputed values
  78. df_imputed = pd.DataFrame(df_imputed, columns=df.columns, index= df.index)
  79. df_imputed*10e6
  81. # %%
  82. #[0,:] all columns in the first row, what type of distribution
  83. df_imputed.iloc[0,:].hist()
  84. # Set the x and y labels
  85. plt.xlabel('Value') # Label for x-axis
  86. plt.ylabel('Frequency') # Label for y-axis
  87. plt.show()
  89. # %%
  90. df_imputed = np.log(df_imputed*10e6) #np.log (...) applies the natural algorithm (log base e) to all values, improve linearity for PCA
  91. #make data more normally distributed and reduce the effects of outliers
  93. # %%
  94. df_imputed.iloc[0,:].hist()
  95. plt.xlabel('Value') # Label for x-axis
  96. plt.ylabel('Frequency') # Label for y-axis
  97. plt.show()
  99. # %%
  100. # Extract group labels from column names
  101. # Jessi code=group_labels = [int(col[-1]) for col in df.columns]
  102. #My group_labels are in the firts row
  103. group_labels = [col[0] for col in df.columns]
  104. group_labels
  106. # %%
  107. #shape of the transposed values
  108. df_imputed.values.T.shape
  110. # %%
  111. # Transpose the DataFrame
  112. df_transposed = df_imputed.T #T stands for transpose it flips the data frame rows into columns and columns into rows
  113. #samples as rows are requiered for PCA and StandardScaler
  115. # %%
  116. # Print the new shape
  117. print(df_transposed.shape)
  119. # %%
  120. df_transposed
  122. # %%
  123. # Standardize features (per column)
  124. from sklearn.preprocessing import StandardScaler
  125. scaler = StandardScaler()
  126. df_scaled = scaler.fit_transform(df_imputed.values) #df_scaled standarized
  128. # %%
  129. # Reconstruct a DataFrame from df_scaled for easier plotting
  130. df_scaled_df = pd.DataFrame(df_scaled, index=df_imputed.index, columns=df_imputed.columns)
  132. # Boxplot per sample (i.e., each box is one sample)
  133. plt.figure(figsize=(14, 6))
  134. df_scaled_df.boxplot(showfliers=False) # Set to True if you want to show outliers
  136. # Plot formatting
  137. plt.title("Boxplot of Standardized Feature Values per Sample", fontsize=16)
  138. plt.xlabel("Samples", fontsize=14)
  139. plt.ylabel("Standardized Feature Values", fontsize=14)
  140. plt.xticks(rotation=90, fontsize=10)
  141. plt.yticks(fontsize=12)
  142. plt.tight_layout()
  143. plt.show()
  145. # %%
  146. print("Shape of df_scaled_df:", df_scaled_df.shape)
  148. # %%
  149. print("Any missing values?", df_scaled_df.isnull().values.any())
  151. # %%
  152. print("Overall mean:", df_scaled_df.values.mean())
  153. print("Overall std:", df_scaled_df.values.std())
  155. # %%
  156. df_scaled_T = df_scaled_df.T # Now shape is (37, 9782)
  158. # %%
  159. print("Shape of df_scaled_df:", df_scaled_df.T.shape)
  161. # %%
  162. print("Any missing values?", df_scaled_df.T.isnull().values.any())
  164. # %%
  165. print("Overall mean:", df_scaled_df.T.values.mean())
  166. print("Overall std:", df_scaled_df.T.values.std())
  168. # %%
  169. # Transpose so samples are rows
  170. df_scaled_T = df_scaled_df.T
  172. # ✅ Convert all column names (features) to strings
  173. df_scaled_T.columns = df_scaled_T.columns.astype(str)
  175. # PCA
  176. from sklearn.decomposition import PCA
  177. pca = PCA(n_components=2)
  178. pca_data = pca.fit_transform(df_scaled_T)
  180. # Step 2: Create DataFrame with PCA results
  181. pca_df = pd.DataFrame(pca_data, columns=["PC1", "PC2"])
  182. pca_df["Sample"] = df_scaled_T.index # Sample names from rows
  183. pca_df["Group"] = pca_df["Sample"].str.split("_").str[0] # Customize based on your naming # not good for
  185. # Step 3: Plot with Matplotlib
  186. fig, ax = plt.subplots(figsize=(10, 8))
  188. # Color by group
  189. for group in pca_df["Group"].unique():
  190. group_data = pca_df[pca_df["Group"] == group]
  191. ax.scatter(group_data["PC1"], group_data["PC2"], label=f"{group}", s=150, alpha=0.7)
  193. # Optional: annotate each sample
  194. for _, row in pca_df.iterrows():
  195. ax.text(row["PC1"] + 0.3, row["PC2"] + 0.3, row["Sample"], fontsize=9, alpha=0.8)
  197. # Variance explained for axis labels
  198. explained_var = PCA(n_components=2).fit(df_scaled_T).explained_variance_ratio_ * 100
  199. ax.set_xlabel(f"PC1 ({explained_var[0]:.1f}%)", fontsize=14)
  200. ax.set_ylabel(f"PC2 ({explained_var[1]:.1f}%)", fontsize=14)
  202. # Final formatting
  203. ax.set_title("PCA Plot (Matplotlib)", fontsize=16, fontweight="bold")
  204. ax.legend(title="Group", fontsize=12)
  205. ax.tick_params(labelsize=12)
  206. plt.tight_layout()
  207. plt.show()
  209. # %%
  210. # Transpose so samples are rows
  211. df_scaled_T = df_scaled_df.T
  213. # ✅ Convert all column names (features) to strings
  214. df_scaled_T.columns = df_scaled_T.columns.astype(str)
  216. # PCA
  217. from sklearn.decomposition import PCA
  218. pca = PCA(n_components=2)
  219. pca_data = pca.fit_transform(df_scaled_T)
  221. # Step 2: Create DataFrame with PCA results
  222. pca_df = pd.DataFrame(pca_data, columns=["PC1", "PC2"])
  223. pca_df["Sample"] = df_scaled_T.index # Sample names from rows
  224. pca_df["Group"] = pca_df["Sample"].str[0] # Customize based on your naming # not good for
  226. # Step 3: Plot with Matplotlib
  227. fig, ax = plt.subplots(figsize=(12, 10))
  229. # Color by group
  230. for group in pca_df["Group"].unique():
  231. group_data = pca_df[pca_df["Group"] == group]
  232. ax.scatter(group_data["PC1"], group_data["PC2"], label=f"{group}", s=300, alpha=0.7) #dot size s=400
  234. # Optional: annotate each sample
  235. for _, row in pca_df.iterrows():
  236. ax.text(row["PC1"] + 0.3, row["PC2"] + 0.3, row["Sample"], fontsize=9, alpha=0.8)
  238. # Variance explained for axis labels
  239. explained_var = PCA(n_components=2).fit(df_scaled_T).explained_variance_ratio_ * 100
  240. ax.set_xlabel(f"PC1 ({explained_var[0]:.1f}%)", fontsize=14)
  241. ax.set_ylabel(f"PC2 ({explained_var[1]:.1f}%)", fontsize=14)
  243. # Final formatting
  244. ax.set_title("PCA Plot (Matplotlib)", fontsize=16, fontweight="bold") #axis label fontsize+
  245. ax.legend(title="Group", fontsize=12)
  246. ax.tick_params(labelsize=14)
  247. plt.tight_layout()
  248. plt.show()
  250. # %%
  252. # Step 1: Perform PCA
  253. pca = PCA(n_components=2)
  254. pca_data = pca.fit_transform(df_scaled_T)
  256. # Step 2: Create DataFrame with PCA results
  257. pca_df = pd.DataFrame(pca_data, columns=["PC1", "PC2"])
  258. pca_df["Sample"] = df_scaled_T.index
  259. pca_df["Group"] = pca_df["Sample"].str[0] # Adjust if needed
  261. # Step 3: Plot
  262. fig, ax = plt.subplots(figsize=(12, 10))
  264. # Plot each group with larger dots
  265. for group in pca_df["Group"].unique():
  266. group_data = pca_df[pca_df["Group"] == group]
  267. ax.scatter(
  268. group_data["PC1"],
  269. group_data["PC2"],
  270. label=f"Group {group}",
  271. s=700,
  272. alpha=0.7
  273. )
  275. # Axis labels with % variance explained
  276. explained = pca.explained_variance_ratio_ * 100
  277. #ax.set_title("PCA Plot", fontsize=20, fontweight='bold')
  278. ax.set_xlabel(f"PC1 ({explained[0]:.1f}%)", fontsize=16)
  279. ax.set_ylabel(f"PC2 ({explained[1]:.1f}%)", fontsize=16)
  281. # Make axis lines (spines) thicker
  282. ax.spines['bottom'].set_linewidth(2)
  283. ax.spines['left'].set_linewidth(2)
  284. ax.spines['top'].set_linewidth(2)
  285. ax.spines['right'].set_linewidth(2)
  287. # Bigger tick labels
  288. ax.tick_params(width=2, length=6, labelsize=25)
  290. # Legend
  291. ax.legend(title="Group", fontsize=12, title_fontsize=14, bbox_to_anchor=(1.05, 1), loc='upper left')
  293. # Tight layout
  294. plt.tight_layout()
  296. # Step 4: Export as SVG
  297. #plt.savefig(r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\PCA_final.svg", format="svg")
  299. # Optional: Show plot
  300. plt.show()
  302. # %%
  303. from sklearn.decomposition import PCA
  304. pca = PCA(n_components=2)
  305. pca_data = pca.fit_transform(df_scaled_T)
  306. explained = pca.explained_variance_ratio_ * 100
  307. print("Explained variance:", pca.explained_variance_ratio_)
  309. # %%
  310. print("Groups detected:", pca_df["Group"].unique())
  311. print("Number of groups:", pca_df["Group"].nunique())
  313. # %%
  314. from sklearn.cluster import KMeans
  316. # Step 1: Cluster PCA data (2D) into 2 clusters
  317. kmeans = KMeans(n_clusters=2, random_state=42)
  318. pca_df["Cluster"] = kmeans.fit_predict(pca_df[["PC1", "PC2"]])
  320. # Step 2: Plot with cluster colors instead of groups
  321. fig, ax = plt.subplots(figsize=(10, 8))
  323. # Define cluster colors
  324. colors = ['red', 'blue'] # Or use more for more clusters
  325. for cluster in pca_df["Cluster"].unique():
  326. cluster_data = pca_df[pca_df["Cluster"] == cluster]
  327. ax.scatter(cluster_data["PC1"], cluster_data["PC2"],
  328. label=f"Cluster {cluster+1}", c=colors[cluster],
  329. s=200, alpha=0.7)
  331. # Axes and legend
  332. explained = pca.explained_variance_ratio_ * 100
  333. ax.set_xlabel(f"PC1 ({explained[0]:.1f}%)", fontsize=14)
  334. ax.set_ylabel(f"PC2 ({explained[1]:.1f}%)", fontsize=14)
  335. ax.set_title("PCA with KMeans Clustering", fontsize=16, fontweight="bold")
  336. ax.legend(title="KMeans Cluster", fontsize=12)
  337. ax.tick_params(labelsize=12)
  339. # Optional: Thicker axes
  340. for spine in ax.spines.values():
  341. spine.set_linewidth(2)
  342. ax.tick_params(width=2, length=6)
  344. plt.tight_layout()
  345. plt.show()
  347. # %%
  348. pd.crosstab(pca_df["Group"], pca_df["Cluster"])
  350. # %%
  351. from sklearn.cluster import KMeans
  352. import matplotlib.pyplot as plt
  354. # Step 1: Cluster PCA data (2 clusters for example)
  355. kmeans = KMeans(n_clusters=2, random_state=42)
  356. pca_df["Cluster"] = kmeans.fit_predict(pca_df[["PC1", "PC2"]])
  358. # Step 2: Get centroids from the KMeans model
  359. centroids = kmeans.cluster_centers_ # shape: (n_clusters, 2)
  361. # Step 3: Plot
  362. fig, ax = plt.subplots(figsize=(10, 8))
  364. # Cluster colors
  365. colors = ['lightblue', 'gray']
  366. for cluster in pca_df["Cluster"].unique():
  367. cluster_data = pca_df[pca_df["Cluster"] == cluster]
  368. ax.scatter(
  369. cluster_data["PC1"],
  370. cluster_data["PC2"],
  371. label=f"Cluster {cluster+1}",
  372. c=colors[cluster],
  373. s=500,
  374. alpha=0.7
  375. )
  377. # Step 4: Plot centroids
  378. ax.scatter(
  379. centroids[:, 0], # PC1
  380. centroids[:, 1], # PC2
  381. marker='X',
  382. s=300,
  383. c='orange',
  384. label='Centroid',
  385. edgecolor='white',
  386. linewidth=2
  387. )
  389. # Axis formatting
  390. explained = pca.explained_variance_ratio_ * 100
  391. ax.set_xlabel(f"PC1 ({explained[0]:.1f}%)", fontsize=22)
  392. ax.set_ylabel(f"PC2 ({explained[1]:.1f}%)", fontsize=22)
  393. ax.set_title("PCA with KMeans Clustering and Centroids", fontsize=18, fontweight="bold")
  394. ax.tick_params(labelsize=22)
  395. for spine in ax.spines.values():
  396. spine.set_linewidth(2)
  397. ax.tick_params(width=2, length=6)
  398. #ax.legend(title="Cluster", fontsize=16)
  399. plt.tight_layout()
  401. #plt.savefig(r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\PCA_withKMeans.svg", format="svg")
  402. plt.show()
  404. # %%
  405. import seaborn as sns
  406. import matplotlib.pyplot as plt
  408. # Step 1: Transpose to get samples as rows
  409. df_plot = df_imputed.T
  411. # Step 2: Add group info based on sample names
  412. df_plot["Group"] = df_plot.index.str[0]
  414. # Step 3: Choose gene of interest (make sure it's in the column names now)
  415. gene_of_interest = "PRPH" # Replace with a valid gene name from df_imputed.index
  417. # Step 4: Move the gene from index to column if necessary
  418. df_plot.columns = df_plot.columns.astype(str)
  419. if gene_of_interest not in df_plot.columns:
  420. df_plot[gene_of_interest] = df_imputed.loc[gene_of_interest].values
  422. # Step 5: Violin plot
  423. plt.figure(figsize=(10, 6))
  424. sns.violinplot(data=df_plot, x="Group", y=gene_of_interest, inner="box", palette="Set2")
  426. # Step 6: Plot styling
  427. plt.title(f"Expression of {gene_of_interest} Across Groups", fontsize=16)
  428. plt.xlabel("Group", fontsize=14)
  429. plt.ylabel("Expression Level", fontsize=14)
  430. plt.xticks(fontsize=12)
  431. plt.yticks(fontsize=12)
  432. plt.tight_layout()
  433. plt.show()
  435. # %%
  436. print(df_imputed.index[:10].tolist())
  438. # %%
  439. import seaborn as sns
  440. import matplotlib.pyplot as plt
  442. # Step 1: Transpose to get samples as rows
  443. df_plot = df_imputed.T
  445. # Step 2: Add group info based on sample names
  446. df_plot["Group"] = df_plot.index.str[0]
  448. # Step 3: Choose gene of interest (make sure it's in the column names now)
  449. gene_of_interest = "SYN1" # Replace with a valid gene name from df_imputed.index
  451. # Step 4: Move the gene from index to column if necessary
  452. df_plot.columns = df_plot.columns.astype(str)
  453. if gene_of_interest not in df_plot.columns:
  454. df_plot[gene_of_interest] = df_imputed.loc[gene_of_interest].values
  456. # Step 5: Violin plot
  457. plt.figure(figsize=(10, 6))
  458. sns.violinplot(data=df_plot, x="Group", y=gene_of_interest, inner="box", palette="Set2")
  460. # Step 6: Plot styling
  461. plt.title(f"Expression of {gene_of_interest} Across Groups", fontsize=16)
  462. plt.xlabel("Group", fontsize=14)
  463. plt.ylabel("Expression Level", fontsize=14)
  464. plt.xticks(fontsize=12)
  465. plt.yticks(fontsize=12)
  466. plt.tight_layout()
  467. plt.show()
  469. # %%
  470. import pandas as pd
  472. # Step 1: Transpose to get samples as rows
  473. df_plot = df_imputed.T.copy() # Now shape: (samples, genes)
  475. # Step 2: Add group info (based on sample name, e.g., 'A1' → group 'A')
  476. df_plot["Group"] = df_plot.index.str[0] # Adjust if your sample names are different
  478. # Step 3: Extract expression of the gene of interest (e.g., TP53)
  479. gene_of_interest = "SLC17A6" # Make sure it's a valid gene in df_imputed.index
  480. gene_data = df_plot[[gene_of_interest, "Group"]] # Keep only gene and group
  482. # Step 4: Export to Excel
  483. output_path = r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\SC17A6_expression_by_group.xlsx"
  484. #gene_data.to_excel(output_path)
  486. print("Exported:", output_path)
  488. # %% [markdown]
  489. # ### Differential expression between clusters.
  491. # %%
  492. # Step 1: Transpose df_imputed to get samples as rows
  493. df_expr = df_imputed.T.copy() # shape: (samples, proteins)
  495. # Step 2: Add Cluster labels to expression matrix
  496. df_expr["Cluster"] = pca_df.set_index("Sample").loc[df_expr.index, "Cluster"]
  498. # Step 3: Group by cluster and compute mean expression
  499. cluster_means = df_expr.groupby("Cluster").mean().T # shape: (proteins, 2 clusters)
  501. # Step 4: Compute absolute difference between clusters
  502. cluster_means["diff"] = abs(cluster_means[0] - cluster_means[1])
  504. # Step 5: Sort by difference and get top 10 proteins
  505. top_proteins = cluster_means.sort_values("diff", ascending=False).head(20)
  506. print("Top 20 proteins separating clusters:\n")
  507. print(top_proteins)
  509. # %%
  510. # Cross-tabulate clusters and original groups
  511. cluster_group_counts = pd.crosstab(pca_df["Cluster"], pca_df["Group"])
  513. # Display the mapping
  514. print(cluster_group_counts)
  516. # %%
  517. cluster_group_counts.T.plot(kind="bar", stacked=True, figsize=(10, 6), colormap="Set2")
  518. plt.title("Group Composition per Cluster")
  519. plt.xlabel("Group")
  520. plt.ylabel("Sample Count")
  521. plt.legend(title="KMeans Cluster")
  522. plt.tight_layout()
  523. plt.show()
  525. # %%
  526. import matplotlib.pyplot as plt
  527. import seaborn as sns
  528. import pandas as pd
  530. # Setup colors and marker shapes
  531. group_colors = {
  532. "A": "forestgreen", "B": "red", "C": "dodgerblue", "D": "darkorange",
  533. "E": "gold", "F": "tomato", "G": "purple", "H": "skyblue"
  534. }
  536. cluster_markers = {
  537. 0: "o", # circle
  538. 1: "s", # square
  539. }
  541. # Create figure
  542. fig, ax = plt.subplots(figsize=(12, 10))
  544. # Plot each sample by group (color) and cluster (shape)
  545. for cluster in sorted(pca_df["Cluster"].unique()):
  546. for group in sorted(pca_df["Group"].unique()):
  547. subset = pca_df[(pca_df["Cluster"] == cluster) & (pca_df["Group"] == group)]
  548. ax.scatter(
  549. subset["PC1"],
  550. subset["PC2"],
  551. color=group_colors.get(group, "gray"),
  552. marker=cluster_markers.get(cluster, "o"),
  553. s=200,
  554. alpha=0.8,
  555. label=f"Group {group}, Cluster {cluster}"
  556. )
  558. # Axis formatting
  559. explained = pca.explained_variance_ratio_ * 100
  560. ax.set_xlabel(f"PC1 ({explained[0]:.1f}%)", fontsize=14)
  561. ax.set_ylabel(f"PC2 ({explained[1]:.1f}%)", fontsize=14)
  562. ax.set_title("PCA: Color by Group, Shape by Cluster", fontsize=18, fontweight="bold")
  563. ax.tick_params(labelsize=12)
  565. # Legend (no duplicates)
  566. handles, labels = ax.get_legend_handles_labels()
  567. unique_labels = dict(zip(labels, handles))
  568. ax.legend(unique_labels.values(), unique_labels.keys(), title="Group + Cluster", bbox_to_anchor=(1.05, 1), loc='upper left')
  570. plt.tight_layout()
  571. plt.show()
  573. # %%
  574. import matplotlib.patches as mpatches
  575. import numpy as np
  577. fig, ax = plt.subplots(figsize=(12, 10))
  579. # Plot each group with unique color, marker shape based on cluster
  580. for cluster in sorted(pca_df["Cluster"].unique()):
  581. for group in sorted(pca_df["Group"].unique()):
  582. subset = pca_df[(pca_df["Group"] == group) & (pca_df["Cluster"] == cluster)]
  583. if subset.empty:
  584. continue
  586. # Plot points
  587. ax.scatter(
  588. subset["PC1"],
  589. subset["PC2"],
  590. color=group_colors.get(group, "gray"),
  591. marker=cluster_markers.get(cluster, "o"),
  592. s=200,
  593. alpha=0.7,
  594. label=f"{group}, Cluster {cluster}"
  595. )
  597. # Add ellipse around the group
  598. if len(subset) >= 3: # Need at least 3 points to define an ellipse
  599. cov = np.cov(subset[["PC1", "PC2"]].T)
  600. lambda_, v = np.linalg.eig(cov)
  601. lambda_ = np.sqrt(lambda_)
  602. ell = mpatches.Ellipse(
  603. xy=(subset["PC1"].mean(), subset["PC2"].mean()),
  604. width=lambda_[0] * 4, # 2 std dev
  605. height=lambda_[1] * 4,
  606. angle=np.rad2deg(np.arccos(v[0, 0])),
  607. edgecolor=group_colors.get(group, "gray"),
  608. facecolor='none',
  609. linewidth=2,
  610. linestyle="--",
  611. alpha=0.7
  612. )
  613. ax.add_patch(ell)
  615. # Axes
  616. explained = pca.explained_variance_ratio_ * 100
  617. ax.set_xlabel(f"PC1 ({explained[0]:.1f}%)", fontsize=14)
  618. ax.set_ylabel(f"PC2 ({explained[1]:.1f}%)", fontsize=14)
  619. ax.set_title("PCA with Ellipses by Group and Clustering Shape", fontsize=18)
  621. # Legend cleanup
  622. handles, labels = ax.get_legend_handles_labels()
  623. unique = dict(zip(labels, handles))
  624. ax.legend(unique.values(), unique.keys(), title="Group + Cluster", bbox_to_anchor=(1.05, 1), loc="upper left")
  626. plt.tight_layout()
  627. plt.show()
  629. # %%
  630. # Step 1: Remove Group A
  631. samples_to_keep = pca_df[pca_df["Group"] != "A"]["Sample"]
  632. df_filtered = df_imputed.loc[:, samples_to_keep]
  634. # Step 2: Transpose and fix column types
  635. df_filtered_T = df_filtered.T.copy()
  636. df_filtered_T.columns = df_filtered_T.columns.astype(str)
  638. # Step 3: Standardize
  639. from sklearn.preprocessing import StandardScaler
  640. scaler = StandardScaler()
  641. df_scaled_filtered = scaler.fit_transform(df_filtered_T)
  643. # ✅ Step 4: Run PCA on filtered data
  644. from sklearn.decomposition import PCA
  645. pca = PCA(n_components=2)
  646. pca_data_filtered = pca.fit_transform(df_scaled_filtered) # 🔥 This was missing
  648. # ✅ Step 5: Run KMeans clustering on PCA output
  649. from sklearn.cluster import KMeans
  650. kmeans = KMeans(n_clusters=2, random_state=42)
  651. clusters = kmeans.fit_predict(pca_data_filtered)
  653. # Step 6: Build new PCA DataFrame
  654. import pandas as pd
  655. pca_filtered_df = pd.DataFrame(pca_data_filtered, columns=["PC1", "PC2"])
  656. pca_filtered_df["Sample"] = df_filtered.columns
  657. pca_filtered_df["Group"] = pca_filtered_df["Sample"].str[0]
  658. pca_filtered_df["Cluster"] = clusters # ✅ Now this works!
  660. # Step 7: Plot
  661. import matplotlib.pyplot as plt
  663. fig, ax = plt.subplots(figsize=(10, 8))
  664. for cluster in pca_filtered_df["Cluster"].unique():
  665. cluster_data = pca_filtered_df[pca_filtered_df["Cluster"] == cluster]
  666. ax.scatter(
  667. cluster_data["PC1"], cluster_data["PC2"],
  668. label=f"Cluster {cluster}", s=200, alpha=0.7
  669. )
  671. explained = pca.explained_variance_ratio_ * 100
  672. ax.set_title("PCA after Removing Group A", fontsize=16)
  673. ax.set_xlabel(f"PC1 ({explained[0]:.1f}%)", fontsize=14)
  674. ax.set_ylabel(f"PC2 ({explained[1]:.1f}%)", fontsize=14)
  675. ax.legend(title="KMeans Cluster")
  676. plt.tight_layout()
  677. plt.show()
  679. # %%
  680. print(f"Explained variance (PC1): {explained[0]:.2f}%")
  681. print(f"Explained variance (PC2): {explained[1]:.2f}%")
  683. # %%
  684. from scipy.stats import zscore
  686. # Step 1: Extract gene expression across all samples
  687. gene = "NES"
  688. gene_expr = df_imputed.loc[gene] # This is a Series: index = sample names, values = expression
  690. # Step 2: Convert to DataFrame and add Cluster info
  691. gene_df = pd.DataFrame({"Expression": gene_expr})
  692. gene_df["Sample"] = gene_df.index
  693. gene_df["Cluster"] = gene_df["Sample"].map(pca_df.set_index("Sample")["Cluster"])
  695. # Step 3: Compute Z-score of expression across all samples
  696. gene_df["Z_score"] = zscore(gene_df["Expression"])
  698. # Step 4: Compare distributions across clusters
  699. import seaborn as sns
  700. import matplotlib.pyplot as plt
  702. plt.figure(figsize=(8, 6))
  703. sns.boxplot(data=gene_df, x="Cluster", y="Z_score", palette="Set2")
  704. sns.stripplot(data=gene_df, x="Cluster", y="Z_score", color="black", alpha=0.6)
  706. plt.title(f"Z-score of {gene} Expression Across Clusters", fontsize=16)
  707. plt.xlabel("KMeans Cluster", fontsize=14)
  708. plt.ylabel("Z-score", fontsize=14)
  709. plt.tight_layout()
  710. plt.show()
  712. # Step 4: Export to Excel
  713. output_path = r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\NES_zscores.xlsx"
  714. #gene_df.to_excel(output_path, index=False)
  716. print("✅ Exported Z-scores to:", output_path)
  718. # %%
  719. from scipy.stats import zscore
  721. # Step 1: Define your list of genes (or use top ones)
  722. genes_of_interest = ["NES", "TUBB3", "GFAP", "PRPH", "SLC17A7", "SLC17A6", "RBFOX3", "MAP2", "SYN1","SOX2","CD44", "SYP", "MKI67", "ALDH1L1", "DLG4", "DCX"] # Replace with your list
  724. # Step 2: Transpose df_imputed to shape (samples x genes)
  725. df_expr = df_imputed.T.copy()
  726. df_expr["Sample"] = df_expr.index
  727. df_expr["Group"] = df_expr["Sample"].str[0] # Adjust if needed
  728. df_expr["Cluster"] = df_expr["Sample"].map(pca_df.set_index("Sample")["Cluster"])
  730. # Step 3: Melt into long format
  731. df_long = df_expr[genes_of_interest + ["Sample", "Group", "Cluster"]].melt(
  732. id_vars=["Sample", "Group", "Cluster"],
  733. var_name="Gene",
  734. value_name="Expression"
  735. )
  737. # Step 4: Compute Z-scores for each gene across all samples
  738. df_long["Z_score"] = df_long.groupby("Gene")["Expression"].transform(zscore)
  740. # Step 5: Export to Excel
  741. output_path = r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\multi_gene_zscores1.xlsx"
  742. #df_long.to_excel(output_path, index=False)
  744. print("✅ Exported multi-gene Z-scores to:", output_path)
  746. # %%
  747. import seaborn as sns
  748. import matplotlib.pyplot as plt
  750. # Use the DataFrame from your earlier export, e.g., `df_long`
  751. plt.figure(figsize=(14, 8))
  752. sns.boxplot(data=df_long, x="Group", y="Z_score", hue="Gene", palette="Set2")
  754. plt.title("Z-score of Multiple Genes by Group", fontsize=16)
  755. plt.xlabel("Group", fontsize=14)
  756. plt.ylabel("Z-score", fontsize=14)
  757. plt.legend(title="Gene", bbox_to_anchor=(1.05, 1), loc="upper left")
  758. plt.tight_layout()
  759. plt.show()
  761. # %%
  762. plt.figure(figsize=(14, 8))
  763. sns.violinplot(
  764. data=df_long,
  765. x="Gene",
  766. y="Z_score",
  767. hue="Cluster", # ✅ Switch hue to cluster instead
  768. palette="Set2",
  769. dodge=True # Separates the clusters side-by-side
  770. )
  772. plt.title("Violin Plot of Gene Z-scores by Cluster", fontsize=16)
  773. plt.xlabel("Gene")
  774. plt.ylabel("Z-score")
  775. plt.legend(title="Cluster", bbox_to_anchor=(1.05, 1), loc="upper left")
  776. plt.tight_layout()
  777. plt.show()
  779. # %%
  780. from scipy.stats import mannwhitneyu
  781. import seaborn as sns
  782. import matplotlib.pyplot as plt
  783. import os
  785. # Create output folder
  786. #output_dir = r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\violin_plots"
  787. #os.makedirs(output_dir, exist_ok=True)
  789. # ✅ Define custom colors (do this outside the loop — one time)
  790. custom_palette = {0: "lightblue", 1: "lightgray"}
  792. # Loop through genes
  793. for gene in df_long["Gene"].unique():
  794. plt.figure(figsize=(6, 5))
  796. # Subset data
  797. df_gene = df_long[df_long["Gene"] == gene]
  798. group0 = df_gene[df_gene["Cluster"] == 0]["Z_score"]
  799. group1 = df_gene[df_gene["Cluster"] == 1]["Z_score"]
  801. # Mann–Whitney U
  802. u_stat, p_val = mannwhitneyu(group0, group1, alternative='two-sided')
  804. # Significance stars
  805. if p_val < 0.001:
  806. sig = "***"
  807. elif p_val < 0.01:
  808. sig = "**"
  809. elif p_val < 0.05:
  810. sig = "*"
  811. else:
  812. sig = "ns"
  814. # ✅ Plot using your custom colors
  815. sns.violinplot(
  816. data=df_gene,
  817. x="Cluster",
  818. y="Z_score",
  819. hue="Cluster",
  820. split=True,
  821. palette=custom_palette,
  822. dodge=False
  823. )
  824. plt.legend().remove()
  826. # Style
  827. plt.title(f"{gene} (Mann–Whitney p = {p_val:.3e}, {sig})", fontsize=16)
  828. plt.xlabel("Cluster", fontsize=16)
  829. plt.ylabel("Z-score", fontsize=16)
  830. plt.xticks(fontsize=16)
  831. plt.yticks(fontsize=16)
  833. # Thicken axes lines
  834. ax = plt.gca()
  835. for spine in ax.spines.values():
  836. spine.set_linewidth(1)
  837. # Major ticks
  838. ax.tick_params(axis='both', which='major', labelsize=16, length=6, width=1.5)
  840. # Add significance annotation
  841. y_max = df_gene["Z_score"].max()
  842. plt.ylim(top=y_max + 1)
  843. plt.plot([0.25, 0.75], [y_max + 0.3]*2, lw=1.5, c='black')
  844. plt.text(0.5, y_max + 0.5, sig, ha='center', va='center', fontsize=16)
  846. # Save as SVG
  847. output_path = os.path.join(output_dir, f"{gene}_violin_cluster.svg")
  848. #plt.savefig(output_path, format="svg", bbox_inches="tight")
  849. plt.tight_layout()
  850. plt.show()
  852. # %% [markdown]
  853. # ## For your volcano plot, you want to compare expression values (not Z-scores) between clusters.
  854. #
  855. # ## So:
  856. # ## ❌ Don't use df_scaled_df or df_scaled_T for volcano
  857. # ## ✅ Use the log-transformed, imputed, normalized data → df_imputed
  859. # %%
  860. # Already log-transformed, normalized, and imputed
  861. df_expression = df_imputed
  863. # %%
  864. df_imputed
  866. # %%
  867. sample_to_group = df_expression.columns.str[0] # assumes sample names start with group letter
  869. # %%
  870. group1_labels = ['A', 'B', 'C', 'D']
  871. group2_labels = ['E', 'F', 'G', 'H']
  873. sample_groups = df_expression.columns.to_series().str[0]
  875. group1_samples = sample_groups[sample_groups.isin(group1_labels)].index
  876. group2_samples = sample_groups[sample_groups.isin(group2_labels)].index
  878. # %%
  879. group1 = df_expression[group1_samples]
  880. group2 = df_expression[group2_samples]
  882. # %%
  883. from scipy.stats import ttest_ind
  884. import numpy as np
  885. import pandas as pd
  886. from statsmodels.stats.multitest import multipletests
  887. import matplotlib.pyplot as plt
  889. # Calculate means
  890. mean_group1 = group1.mean(axis=1)
  891. mean_group2 = group2.mean(axis=1)
  893. # Log2 fold change (already log-transformed)
  894. log2fc = mean_group2 - mean_group1
  896. # t-test
  897. pvals = ttest_ind(group2.T, group1.T, axis=0, equal_var=False).pvalue
  899. # FDR correction (Benjamini-Hochberg)
  900. rejected, pvals_corrected, _, _ = multipletests(pvals, alpha=0.05, method='fdr_bh')
  902. # Create volcano dataframe
  903. volcano_df = pd.DataFrame({
  904. "Gene": df_expression.index,
  905. "log2FC": log2fc,
  906. "pval": pvals,
  907. "pval_adj": pvals_corrected,
  908. "significant": rejected
  909. })
  910. volcano_df["neg_log10_pval"] = -np.log10(volcano_df["pval_adj"].clip(lower=1e-300))
  911. volcano_df = volcano_df.replace([np.inf, -np.inf], np.nan).dropna()
  913. # Volcano plot then I
  914. plt.figure(figsize=(10, 6))
  915. plt.scatter(
  916. volcano_df["log2FC"],
  917. volcano_df["neg_log10_pval"],
  918. c=volcano_df["significant"].map({True: "red", False: "gray"}),
  919. alpha=0.7
  920. )
  922. # Label all genes with abs(log2FC) > 1 and FDR < 0.05
  923. highlighted = volcano_df[(volcano_df["significant"]) & (volcano_df["log2FC"].abs() > 1)]
  925. for _, row in highlighted.iterrows():
  926. plt.text(row["log2FC"], row["neg_log10_pval"], row["Gene"],
  927. fontsize=5, ha='center', va='bottom')
  929. # Threshold lines
  930. plt.axhline(-np.log10(0.05), linestyle='--', color='gray')
  931. plt.axvline(1, linestyle='--', color='gray')
  932. plt.axvline(-1, linestyle='--', color='gray')
  934. plt.title("Volcano Plot with FDR Correction")
  935. plt.xlabel("Log₂ Fold Change")
  936. plt.ylabel("-Log₁₀ Adjusted p-value")
  937. plt.tight_layout()
  938. plt.show()
  940. # %% [markdown]
  941. # ## log2fc = mean_group2 - mean_group1
  942. # ##So:
  943. #
  944. # ##If log2FC > 0, the gene is higher in Group 2
  945. #
  946. # ##If log2FC < 0, it’s higher in Group 1
  948. # %%
  949. sig_genes_df = volcano_df[
  950. (volcano_df["pval_adj"] < 0.05) &
  951. (volcano_df["log2FC"].abs() > 1)
  952. ]
  954. # Sort by most significant
  955. sig_genes_df = sig_genes_df.sort_values("pval_adj")
  957. # View top 10
  958. print(sig_genes_df.head(10))
  960. # Export all to Excel
  961. sig_genes_df.to_excel("significant_genes_filtered.xlsx", index=False)
  963. # %%
  964. # Calculate means
  965. mean_group1 = group1.mean(axis=1)
  966. mean_group2 = group2.mean(axis=1)
  968. # Log2 fold change (already log-transformed)
  969. log2fc = mean_group2 - mean_group1
  971. # t-test
  972. pvals = ttest_ind(group2.T, group1.T, axis=0, equal_var=False).pvalue
  974. # FDR correction (Benjamini-Hochberg)
  975. rejected, pvals_corrected, _, _ = multipletests(pvals, alpha=0.05, method='fdr_bh')
  977. # Create volcano dataframe
  978. volcano_df = pd.DataFrame({
  979. "Gene": df_expression.index,
  980. "log2FC": log2fc,
  981. "pval": pvals,
  982. "pval_adj": pvals_corrected,
  983. "significant": rejected
  984. })
  985. volcano_df["neg_log10_pval"] = -np.log10(volcano_df["pval_adj"].clip(lower=1e-300))
  986. volcano_df = volcano_df.replace([np.inf, -np.inf], np.nan).dropna()
  988. # Volcano plot
  989. plt.figure(figsize=(10, 6))
  990. plt.scatter(
  991. volcano_df["log2FC"],
  992. volcano_df["neg_log10_pval"],
  993. c=volcano_df["significant"].map({True: "lightblue", False: "gray"}),
  994. alpha=0.7
  995. )
  997. # 🎯 Annotate specific genes of interest
  998. genes_to_annotate = ["SOX2", "NES", "MIK67", "POU3F", "NOTCH1", "SLC6A9", "GFAP", "GRM2", "GRM5", "RGS10", "FOXP1", "CD44", "ALDH1L1", "SYT7", "GLUL", "SYT11", "SCGN", "SLC6A1", "SLC6A9" "NOTCH1", "DCX", "SDC1", "EGFR", "PHPR"] # ← customize this
  1000. # 1. Not significant
  1001. df_gray = volcano_df[~volcano_df["significant"]]
  1003. # 2. Significant but not in your gene list
  1004. df_lightblue = volcano_df[
  1005. (volcano_df["significant"]) & (~volcano_df["Gene"].isin(genes_to_annotate))
  1006. ]
  1008. # 3. Highlighted genes (e.g., NES, TP53, etc.)
  1009. df_pink = volcano_df[volcano_df["Gene"].isin(genes_to_annotate)]
  1011. plt.figure(figsize=(10, 6))
  1012. ax = plt.gca()
  1014. # Plot gray first
  1015. ax.scatter(df_gray["log2FC"], df_gray["neg_log10_pval"],
  1016. color="gray", alpha=0.6, s=30, label="Not significant")
  1018. # Plot lightblue next
  1019. ax.scatter(df_lightblue["log2FC"], df_lightblue["neg_log10_pval"],
  1020. color="lightblue", alpha=0.7, s=30, label="Significant")
  1022. # Plot pink last (so it’s on top)
  1023. ax.scatter(df_pink["log2FC"], df_pink["neg_log10_pval"],
  1024. color="bisque", edgecolor="darkorange", linewidth=1.0,
  1025. s=60, label="Highlighted Genes")
  1027. for _, row in df_pink.iterrows():
  1028. ax.text(row["log2FC"], row["neg_log10_pval"], row["Gene"],
  1029. fontsize=10, ha='center', va='bottom', fontweight='bold', color='black')
  1030. plt.axhline(-np.log10(0.05), linestyle='--', color='gray')
  1031. plt.axvline(1, linestyle='--', color='gray')
  1032. plt.axvline(-1, linestyle='--', color='gray')
  1034. plt.xlabel("Log₂ Fold Change", fontsize=16)
  1035. plt.ylabel("−Log₁₀ Adjusted p-value", fontsize=16)
  1036. plt.title("Volcano Plot with Layered Highlighting", fontsize=16)
  1037. #plt.legend(frameon=True)
  1038. plt.tight_layout()
  1039. #plt.savefig(r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\Volcano plot.svg", format="svg")
  1040. plt.show()
  1042. # %%
  1043. # 1. Filter volcano_df for significant DE genes
  1044. sig_genes_df = volcano_df[
  1045. (volcano_df["pval_adj"] < 0.05) &
  1046. (volcano_df["log2FC"].abs() > 1)
  1047. ].copy()
  1049. # 2. Sort by adjusted p-value
  1050. sig_genes_df = sig_genes_df.sort_values("pval_adj")
  1052. # 3. (Optional) View top 10
  1053. print("Top significant genes:\n", sig_genes_df.head(10))
  1055. # 4. Export to Excel
  1056. output_path = r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\sig_genes.xlsx"
  1057. #sig_genes_df.to_excel(output_path, index=False)
  1058. print(f"Exported to: {output_path}")
  1060. # %%
  1061. #!pip install gprofiler-official
  1063. # %%
  1064. from gprofiler import GProfiler
  1066. # Get your gene list
  1067. gene_list = sig_genes_df["Gene"].tolist()
  1069. # Create gProfiler object
  1070. gp = GProfiler(return_dataframe=True)
  1072. # Run GO/pathway enrichment
  1073. results = gp.profile(
  1074. organism="hsapiens",
  1075. query=gene_list,
  1076. user_threshold=0.05,
  1077. sources=["GO:BP", "GO:MF", "GO:CC", "KEGG", "REAC", "WP"]
  1078. )
  1080. # Step 4: Check results
  1081. print(f"Submitted {len(gene_list)} genes.")
  1083. if results.empty:
  1084. print("⚠️ No significant enrichment found. Try relaxing thresholds or using more genes.")
  1085. else:
  1086. print("✅ Enrichment completed. Columns available:", results.columns.tolist())
  1088. # Only show if 'term_name' exists
  1089. if "term_name" in results.columns:
  1090. print(results[["term_name", "p_value", "source"]].head(10))
  1091. else:
  1092. print("⚠️ 'term_name' column not found in results. Full results:")
  1093. print(results.head())
  1094. # Save results
  1095. output_path = r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\gprofiler_enrichment_results.xlsx"
  1096. #results.to_excel(output_path, index=False)
  1097. print(f"✅ Saved results to {output_path}")
  1099. # %%
  1100. print("Columns in results:", results.columns.tolist())
  1101. print(results.head())
  1103. # %%
  1104. print(f"Number of enriched terms found: {results.shape[0]}")
  1106. # %%
  1107. import seaborn as sns
  1108. import matplotlib.pyplot as plt
  1110. # Make sure results aren't empty
  1111. if not results.empty and "name" in results.columns:
  1112. # Step 1: Pick top N terms (you can change 10)
  1113. top_terms = results.sort_values("p_value").head(10)
  1115. # Step 2: Barplot of top pathways
  1116. plt.figure(figsize=(8, 6))
  1117. sns.barplot(
  1118. y="name",
  1119. x="-log10(p_value)",
  1120. data=top_terms.assign(**{"-log10(p_value)": -np.log10(top_terms["p_value"])}),
  1121. palette="viridis"
  1122. )
  1123. plt.title("Top 10 Enriched Pathways (g:Profiler)", fontsize=16)
  1124. plt.xlabel("-Log₁₀ p-value", fontsize=14)
  1125. plt.ylabel("Pathway", fontsize=14)
  1126. plt.tight_layout()
  1127. plt.show()
  1128. else:
  1129. print("⚠️ No results to plot.")
  1131. # %%
  1132. # Find the row for the pathway you care about
  1133. row = results[results["name"] == "cell junction"]
  1135. if not row.empty:
  1136. print("Matched Genes:", row["intersection_size"].values[0])
  1137. else:
  1138. print("⚠️ No pathway found with that exact name!")
  1140. # Show the genes (usually in 'intersection')
  1141. print(row["intersection_size"].values[0])
  1143. # %%
  1144. import seaborn as sns
  1145. import matplotlib.pyplot as plt
  1146. import numpy as np
  1148. # Confirm 'results' is populated and has expected columns
  1149. if not results.empty and "name" in results.columns and "p_value" in results.columns:
  1150. # Step 1: Compute -log10(p-value)
  1151. results["log_p"] = -np.log10(results["p_value"])
  1153. # Step 2: Get top 10 terms by significance
  1154. top_terms = results.nsmallest(10, "p_value")
  1156. # Step 3: Bubble plot
  1157. plt.figure(figsize=(10, 8))
  1158. sns.scatterplot(
  1159. data=top_terms,
  1160. x="intersection_size", # Number of DEGs in each term
  1161. y="name", # GO term or pathway name
  1162. size="intersection_size",
  1163. hue="log_p",
  1164. sizes=(100, 1000),
  1165. palette="viridis",
  1166. legend="brief"
  1167. )
  1169. plt.xlabel("Gene Count in Term", fontsize=12)
  1170. plt.ylabel("Pathway / GO Term", fontsize=12)
  1171. plt.title("Top Enriched Terms (Bubble Plot)", fontsize=14)
  1172. plt.tight_layout()
  1173. plt.show()
  1174. else:
  1175. print("⚠️ No valid enrichment results found to plot.")
  1177. # %%
  1178. import matplotlib.pyplot as plt
  1179. import seaborn as sns
  1180. import numpy as np
  1182. # List of GO terms or pathway names you're interested in (case-insensitive match)
  1183. terms_of_interest = [
  1184. "cell differentiation",
  1185. "astrocyte differentiation",
  1186. "nervous system development",
  1187. "neurogenesis",
  1188. "generation of neurons",
  1189. "neuron differentiation",
  1190. "cortical cytoskeleton organization",
  1191. "synapse",
  1192. "axon development",
  1193. "gliogenesis",
  1194. "Neuroinflammation and glutamatergic signaling",
  1195. ]
  1197. # --- 1. Prepare p-value for color scale ---
  1198. filtered_terms["p_value_capped"] = filtered_terms["p_value"].clip(lower=1e-10) # avoid log(0)
  1199. filtered_terms["p_label"] = filtered_terms["p_value"].apply(lambda x: f"{x:.1e}")
  1202. # Sort to control y-axis order (optional)
  1203. filtered_terms = filtered_terms.sort_values("intersection_size", ascending=True).reset_index(drop=True)
  1205. # --- Calculate plot boundaries safely ---
  1206. max_x = filtered_terms["intersection_size"].max()
  1207. x_offset = max_x * 0.40 # shift p-values to the right by ~10%
  1208. x_buffer = max_x * 0.25 # add buffer to x-axis so text isn't cut off
  1210. # --- Set up plot ---
  1211. fig, ax = plt.subplots(figsize=(12, 8))
  1214. bubble = sns.scatterplot(
  1215. data=filtered_terms,
  1216. x="intersection_size",
  1217. y="name",
  1218. size="intersection_size",
  1219. hue="p_value",
  1220. palette="Blues_r",
  1221. sizes=(200, 1200),
  1222. alpha=0.8,
  1223. edgecolor="gray",
  1224. legend="brief",
  1225. ax=ax
  1226. )
  1228. # --- Add p-value text BELOW bubble ---
  1229. for i, row in filtered_terms.iterrows():
  1230. ax.text(
  1231. row["intersection_size"],
  1232. i - 0.5, # further down
  1233. row["p_label"],
  1234. ha="left",
  1235. va="top",
  1236. fontsize=12,
  1237. color="black"
  1238. )
  1240. # --- Adjust font sizes ---
  1241. ax.set_yticks(range(len(filtered_terms)))
  1242. ax.set_yticklabels(filtered_terms["name"], fontsize=14)
  1243. ax.tick_params(axis='x', labelsize=12)
  1245. ax.set_xlabel("Gene Count in Term", fontsize=14)
  1246. ax.set_ylabel("GO Term", fontsize=14)
  1247. ax.set_title("Selected GO Terms (Enrichment Bubble Plot)", fontsize=16)
  1249. # --- Prevent clipping of bubbles or labels ---
  1250. ax.set_xlim(0, max_x + x_buffer)
  1251. ax.set_ylim(-1, len(filtered_terms) - 0.5)
  1253. # --- Fix legend position outside plot ---
  1254. sns.move_legend(bubble, "upper left", bbox_to_anchor=(1.02, 1), frameon=True)
  1256. # --- Improve layout ---
  1257. plt.grid(axis="x", linestyle="--", alpha=0.3)
  1258. plt.tight_layout(rect=[0, 0, 1, 1]) # leave space on right for legend
  1260. #plt.savefig(r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\Enrichment Bubble Plot.svg", format="svg", bbox_inches="tight")
  1261. plt.show()
  1263. # %%
  1264. # Example: Plot a single protein across conditions
  1265. proteins_of_interest = ["NES", "MAP2", "GFAP"] # Replace with your protein
  1267. # Make sure your expression data is in long format
  1269. # Melt data for seaborn plotting
  1270. expr_long = df_imputed.loc[proteins_of_interest].T.reset_index().melt(id_vars="index", var_name="Protein", value_name="Expression")
  1271. expr_long.rename(columns={"index": "Sample"}, inplace=True)
  1273. # Add group info from sample names if encoded
  1274. expr_long["Group"] = expr_long["Sample"].str[0] # Adjust if needed
  1276. # Plot
  1277. plt.figure(figsize=(12, 6))
  1278. sns.boxplot(data=expr_long, x="Protein", y="Expression", hue="Group")
  1279. sns.stripplot(data=expr_long, x="Protein", y="Expression", hue="Group", dodge=True, color='black', size=4, alpha=0.5)
  1280. plt.title("Expression of Selected Proteins by Group")
  1281. plt.xticks(rotation=45)
  1282. plt.tight_layout()
  1283. plt.show()
  1285. # %% [markdown]
  1286. # # df_expression = df_imputed # ⬅️ This was your imputed expression matrix df_imputed = df_expression # If df_expression is already defined
  1288. # %%
  1289. df_imputed = df_expression # If df_expression is already defined
  1291. # %%
  1292. df
  1294. # %%
  1295. # Assuming your DataFrame is named 'df'
  1296. column_sums = df.sum(axis=0) # Calculate the sum of each column
  1298. # Divide each element by the corresponding column sum
  1299. df_normalized = df.div(column_sums, axis=1)
  1300. df_normalized
  1302. # %%
  1303. # Handle missing values (e.g., impute with mean or median), imputation may contain very small numbers
  1304. # Create a KNN imputer object
  1305. imputer = KNNImputer(n_neighbors=5) # Adjust n_neighbors as needed
  1307. # Impute missing values using KNN
  1308. df_imputed = imputer.fit_transform(df_normalized)
  1310. # Create a new DataFrame with imputed values
  1311. df_imputed = pd.DataFrame(df_imputed, columns=df.columns, index= df.index)
  1312. df_imputed*10e6
  1314. # %%
  1315. import pandas as pd
  1316. import numpy as np
  1317. import matplotlib.pyplot as plt
  1318. import seaborn as sns
  1319. from scipy.stats import mannwhitneyu
  1320. import os
  1321. import math
  1323. # --- Scale expression data ---
  1324. df_scaled = df_imputed * 1e7
  1326. # --- Assign groups from sample names ---
  1327. group1_labels = ['A', 'B', 'C', 'D']
  1328. group2_labels = ['E', 'F', 'G', 'H']
  1329. sample_to_group = df_scaled.columns.str[0]
  1330. group_map = sample_to_group.map(lambda x: 'Group 1' if x in group1_labels else ('Group 2' if x in group2_labels else 'Other'))
  1332. # --- Convert to long-form dataframe ---
  1333. df_long = df_scaled.T.copy()
  1334. df_long["Group"] = group_map.values
  1335. df_long = df_long.reset_index().melt(id_vars=["index", "Group"], var_name="Gene", value_name="Expression")
  1336. df_long.rename(columns={"index": "Sample"}, inplace=True)
  1338. # --- Genes of interest ---
  1339. genes_of_interest = ["STAT3", "ZEB2", "GFAP", "PLPP3", "SOX9", "NR3C1"] # add more as needed
  1340. n_genes = len(genes_of_interest)
  1342. # --- Grid dimensions (2 columns) ---
  1343. n_cols = 2
  1344. n_rows = math.ceil(n_genes / n_cols)
  1346. # --- Set up the figure ---
  1347. fig, axes = plt.subplots(n_rows, n_cols, figsize=(14, 5 * n_rows), squeeze=False)
  1349. # --- Plot each gene in grid ---
  1350. for idx, gene in enumerate(genes_of_interest):
  1351. row = idx // n_cols
  1352. col = idx % n_cols
  1353. ax = axes[row][col]
  1355. df_gene = df_long[df_long["Gene"] == gene]
  1356. group1 = df_gene[df_gene["Group"] == "Group 1"]["Expression"]
  1357. group2 = df_gene[df_gene["Group"] == "Group 2"]["Expression"]
  1359. # Mann–Whitney U test
  1360. u_stat, p_val = mannwhitneyu(group1, group2, alternative="two-sided")
  1361. sig = "***" if p_val < 0.001 else "**" if p_val < 0.01 else "*" if p_val < 0.05 else "ns"
  1363. # Plot
  1364. sns.violinplot(data=df_gene, x="Group", y="Expression", palette=["lightblue", "lightgray"], ax=ax, inner=None)
  1365. sns.stripplot(data=df_gene, x="Group", y="Expression", color="black", size=4, jitter=True, alpha=0.6, ax=ax)
  1367. # Significance annotation
  1368. y_max = df_gene["Expression"].max()
  1369. ax.set_ylim(top=y_max + 0.25 * y_max)
  1370. ax.plot([0, 1], [y_max + 0.05 * y_max] * 2, lw=1.5, c="black")
  1371. ax.text(0.5, y_max + 0.1 * y_max, sig, ha="center", va="center", fontsize=14)
  1373. # Style
  1374. ax.set_title(f"{gene} (p = {p_val:.2e}, {sig})", fontsize=13)
  1375. ax.set_xlabel("")
  1376. ax.set_ylabel("Expression", fontsize=12)
  1377. ax.tick_params(axis="both", labelsize=11)
  1379. # --- Hide unused subplots if total isn't a full grid ---
  1380. for i in range(n_genes, n_rows * n_cols):
  1381. fig.delaxes(axes[i // n_cols][i % n_cols])
  1383. # --- Final layout and export ---
  1384. plt.tight_layout()
  1385. output_path = r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\violin_plots\multi_gene_violin_grid.svg"
  1386. #plt.savefig(output_path, format="svg", bbox_inches="tight")
  1387. plt.show()
  1389. # %%
  1390. import numpy as np
  1391. import pandas as pd
  1392. import matplotlib.pyplot as plt
  1393. import seaborn as sns
  1394. from scipy.stats import mannwhitneyu
  1396. # --- List of genes to analyze ---
  1397. genes_of_interest = ["STAT3", "ZEB2", "GFAP", "SOX9", "NR3C1", "NOTCH1", "QKI", "SYN1", "DLG4", "SLC17A7", "SLC17A6", "CCNE1", "CCNA2", "DCX", "EGFR", "SDC1", "SOX2", "PRPH"] # customize as needed
  1399. # --- Setup for results ---
  1400. results = []
  1402. for gene in genes_of_interest:
  1403. group1 = df_scaled.loc[gene, df_scaled.columns.str[0].isin(['A', 'B', 'C', 'D'])]
  1404. group2 = df_scaled.loc[gene, df_scaled.columns.str[0].isin(['E', 'F', 'G', 'H'])]
  1406. # Compute log2 fold change
  1407. fc = group1.mean() / group2.mean()
  1408. log2fc = np.log2(fc)
  1410. # Mann–Whitney U test
  1411. u_stat, p_val = mannwhitneyu(group1, group2, alternative='two-sided')
  1413. # Significance label
  1414. if p_val < 0.001:
  1415. sig = '***'
  1416. elif p_val < 0.01:
  1417. sig = '**'
  1418. elif p_val < 0.05:
  1419. sig = '*'
  1420. else:
  1421. sig = 'ns'
  1423. # Collect
  1424. results.append({"Gene": gene, "log2FC": log2fc, "pval": p_val, "sig": sig})
  1426. # --- Convert to DataFrame ---
  1427. fc_df = pd.DataFrame(results)
  1429. # --- Plot ---
  1430. plt.figure(figsize=(10, 6))
  1431. sns.barplot(data=fc_df, x="Gene", y="log2FC", color="lightblue", edgecolor="gray")
  1433. # Add significance stars above bars
  1434. for i, row in fc_df.iterrows():
  1435. y = row["log2FC"]
  1436. offset = 0.05 * np.max(np.abs(fc_df["log2FC"]))
  1437. plt.text(i, y + np.sign(y) * offset, row["sig"], ha='center', va='bottom' if y >= 0 else 'top', fontsize=14)
  1439. # Style
  1440. plt.axhline(0, color="black", lw=1)
  1441. plt.ylabel("log₂ Fold Change (Group1 / Group2)", fontsize=13)
  1442. plt.xlabel("")
  1443. plt.title("Fold Change Between Groups with Expression Significance", fontsize=14)
  1444. plt.xticks(fontsize=12)
  1445. plt.yticks(fontsize=12)
  1446. plt.tight_layout()
  1448. # Expand y-limits so stars fit
  1449. y_max = fc_df["log2FC"].max()
  1450. y_min = fc_df["log2FC"].min()
  1451. y_range = y_max - y_min
  1452. plt.ylim(y_min - 0.2 * y_range, y_max + 0.2 * y_range)
  1454. output_path = r"C:\Users\MunozestraJ\OneDrive - Cedars-Sinai Health System\Proteomics\Dox experiment\foldchange astrocytes differentiation.svg"
  1455. #plt.savefig(output_path, format="svg", bbox_inches="tight")
  1457. plt.show()
  1459. # %%
  1460. # Display sorted fold change table
  1461. print(fc_df.sort_values("log2FC", ascending=False))
  1463. # %%
  1464. import numpy as np
  1465. import pandas as pd
  1466. import matplotlib.pyplot as plt
  1467. import seaborn as sns
  1468. from scipy.stats import mannwhitneyu
  1470. # --- List of genes to analyze ---
  1471. #genes_of_interest = ["SLC6A9", "GRM2", "PSAT1", "SHMT1", "GLUL", "GRM5", "SLC1A3", "GRIK1"] # customize as needed
  1472. genes_of_interest = ["STAT3", "ZEB2", "GFAP", "SOX9", "NR3C1", "NOTCH1", "QKI"]
  1473. # --- Setup for results ---
  1474. results = []
  1476. for gene in genes_of_interest:
  1477. group1 = df_scaled.loc[gene, df_scaled.columns.str[0].isin(['A', 'B', 'C', 'D'])]
  1478. group2 = df_scaled.loc[gene, df_scaled.columns.str[0].isin(['E', 'F', 'G', 'H'])]
  1480. # Compute log2 fold change
  1481. fc = group2.mean() / group1.mean()
  1482. log2fc = np.log2(fc)
  1484. # Mann–Whitney U test
  1485. u_stat, p_val = mannwhitneyu(group1, group2, alternative='two-sided')
  1487. # Significance label
  1488. if p_val < 0.001:
  1489. sig = '***'
  1490. elif p_val < 0.01:
  1491. sig = '**'
  1492. elif p_val < 0.05:
  1493. sig = '*'
  1494. else:
  1495. sig = 'ns'
  1497. # Collect
  1498. results.append({"Gene": gene, "log2FC": log2fc, "pval": p_val, "sig": sig})
  1500. # --- Convert to DataFrame ---
  1501. fc_df = pd.DataFrame(results)
  1503. # --- Prepare plot ---
  1504. plt.figure(figsize=(10, 6))
  1506. # Sort genes by fold change for visual clarity
  1507. fc_df_sorted = fc_df.sort_values("log2FC", ascending=False).reset_index(drop=True)
  1509. # Plot dots (log2FC on y-axis)
  1510. sns.stripplot(
  1511. data=fc_df_sorted,
  1512. x="log2FC",
  1513. y="Gene",
  1514. size=10,
  1515. color="lightblue",
  1516. edgecolor="orange",
  1517. )
  1519. # Add vertical line at 0 (no change)
  1520. plt.axvline(0, color="black", lw=1)
  1522. # Add significance stars next to dots
  1523. x_range = fc_df_sorted["log2FC"].max() - fc_df_sorted["log2FC"].min()
  1524. x_offset = 0.03 * x_range # offset stars slightly to the right
  1526. for i, row in fc_df_sorted.iterrows():
  1527. x = row["log2FC"]
  1528. y = i
  1529. plt.text(
  1530. x + x_offset,
  1531. y,
  1532. row["sig"],
  1533. va="center",
  1534. ha="left",
  1535. fontsize=13,
  1536. color="black"
  1537. )
  1539. # Style
  1540. plt.xlabel("log₂ Fold Change (Group1 / Group2)", fontsize=13)
  1541. plt.ylabel("")
  1542. plt.title("Fold Change by Gene with Significance", fontsize=14)
  1543. plt.xticks(fontsize=12)
  1544. plt.yticks(fontsize=12)
  1545. plt.tight_layout()
  1546. plt.show()
  1548. # %%
  1549. import pandas as pd
  1550. import seaborn as sns
  1551. import matplotlib.pyplot as plt
  1552. from scipy.stats import ttest_ind
  1554. # --- Protein list to plot ---
  1555. proteins_of_interest = ["STAT3", "ZEB2", "GFAP", "PLPP3", "SOX9", "NR3C1"] # Update as needed
  1557. # --- Melt expression data ---
  1558. expr_long = df_imputed.loc[proteins_of_interest].T.reset_index().melt(
  1559. id_vars="index", var_name="Protein", value_name="Expression"
  1560. )
  1561. expr_long.rename(columns={"index": "Sample"}, inplace=True)
  1563. # --- Assign groups using first letter of sample name ---
  1564. expr_long["Group"] = expr_long["Sample"].str[0]
  1566. # --- Plot: One violin plot per protein ---
  1567. plt.figure(figsize=(12, 6 * len(proteins_of_interest)))
  1569. for i, protein in enumerate(proteins_of_interest):
  1570. plt.subplot(len(proteins_of_interest), 1, i + 1)
  1572. # Subset data for this protein
  1573. data = expr_long[expr_long["Protein"] == protein]
  1575. # Group split
  1576. group_values = [g["Expression"].values for _, g in data.groupby("Group")]
  1577. if len(group_values) == 2:
  1578. stat, pval = ttest_ind(*group_values)
  1579. else:
  1580. pval = float('nan')
  1582. # Determine significance label
  1583. if pval < 0.001:
  1584. sig_label = "***"
  1585. elif pval < 0.01:
  1586. sig_label = "**"
  1587. elif pval < 0.05:
  1588. sig_label = "*"
  1589. else:
  1590. sig_label = "ns"
  1592. # Plot violin and strip
  1593. sns.violinplot(data=data, x="Group", y="Expression", inner=None, palette="pastel")
  1594. sns.stripplot(data=data, x="Group", y="Expression", color="black", size=4, alpha=0.6)
  1596. # Annotate significance
  1597. y_max = data["Expression"].max()
  1598. y_min = data["Expression"].min()
  1599. y_range = y_max - y_min
  1600. plt.text(0.5, y_max + 0.05 * y_range, sig_label, ha="center", fontsize=14)
  1602. # Titles and labels
  1603. plt.title(f"{protein} (p = {pval:.2e})")
  1604. plt.xlabel("Condition")
  1605. plt.ylabel("Expression")
  1607. plt.tight_layout()
  1608. plt.show()
  1610. # %%
  1613. # %%
  1616. # %%
  1619. # %%

PCA and Volcano plot.ipynb at commit 9be7e24, no license · at the source

Overview

Authors: Jesús Muñoz-Estrada1, Andrew Mostafania1, Lahiruni Halwatura1, Ali Haghani1, Yuming Jiang1, Jesse G Meyer1
ORCID iDs: Jesse G Meyer
  1. Department of Computational Biomedicine, Smidt Heart Institute, Cedars Sinai Medical Center, Los Angeles, California, USA
Institutions: Cedars-Sinai Medical Center (United States); Cedars-Sinai Smidt Heart Institute (United States)
Journal: Molecular & cellular proteomics : MCP, volume 25, issue 7, article 101604
Dates: received 7 July 2025; published online 17 June 2026; in print July 2026
Type: Research article · Language: English
License: CC BY-NC-ND
Identifiers: DOI 10.1016/j.mcpro.2026.101604 · PMID 42309337 · PMCID PMC13382583 · OpenAlex W7164999840
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: genetics / omics (modality), human (organism), cellular / molecular (subfield)
Methods: Preprocessing, Statistics, Smoothing, state filtering, decompositions, Connectivity, Machine learning
Keywords: KOLF2.1J, induced pluripotent stem cells (iPSCs), Neurogenin-2 (NGN2) stable integration, AAVS1, EF1α transgene silencing, single-cell proteomics, bulk proteomics, iPSC-derived neuronal-enriched cultures, Notch signaling, NGN2 dosage
MeSH: Basic Helix-Loop-Helix Proteins*, Gene Dosage*, Induced Pluripotent Stem Cells*, Nerve Tissue Proteins*, Neurons*, Animals, Cell Differentiation, Cells, Cultured, DNA Methylation, Humans, Promoter Regions, Genetic, Proteomics, Receptors, Notch (* major topic)
Topic: Peptidase Inhibition and Analysis (Oncology, Medicine), according to OpenAlex
Funding: NIA NIH HHS (R21 AG074234); National Institute on Aging; NIGMS NIH HHS (R35 GM142502); National Institute of General Medical Sciences
Citations: not cited yet (Europe PMC); 37 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 7 matches between paragraphs and lines of code.

xomicsdatascience/generation_of_iNeurons

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: 9be7e24aeb13dcdd0470bc641b06cf1d171ab335, 30 January 2026
Languages: Jupyter (10)
Size: 18 files, 10 scripts
Software Heritage: not archived
Found in: “Code Availability”
Holds: README, 10 notebooks
Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: NumPy (10 files), pandas (10 files), Matplotlib (9 files), seaborn (9 files), SciPy (6 files), scikit-learn (4 files), statsmodels (2 files), anndata (1 file), Plotly (1 file), Scanpy (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
11 files
At the source: github.com/xomicsdatascience/generation_of_iNeurons

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/10.1016/j.mcpro.2026.101604.

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;
  • 10 scripts, each with its path and the digest of its content;
  • 7 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.

Data availability statement

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

  • no repository, dataset or request procedure was recognized in it

Read it in the paper: doi.org/10.1016/j.mcpro.2026.101604.

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 2, 28 September 2026

  • Authors: added Jesse G Meyer (0000-0003-2753-3926); removed Jesse G Meyer

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 6 authors, 10 keywords, 13 MeSH terms, 4 funders, 37 references.

Cite

This paper

Muñoz-Estrada, J., Mostafania, A., Halwatura, L., Haghani, A., Jiang, Y., & Meyer, J. G. (2026). Optimizing NGN2 Dosage Enhances the Neuronal Enrichment of iPSC-Derived Neuronal Cultures. Molecular & cellular proteomics : MCP, 25(7), 101604. https://doi.org/10.1016/j.mcpro.2026.101604

BibTeX

@article{munozestrada2026optimizing,
author = {Muñoz-Estrada, Jesús and Mostafania, Andrew and Halwatura, Lahiruni and Haghani, Ali and Jiang, Yuming and Meyer, Jesse G},
title = {{Optimizing NGN2 Dosage Enhances the Neuronal Enrichment of iPSC-Derived Neuronal Cultures}},
journal = {Molecular \& cellular proteomics : MCP},
year = {2026},
month = jun,
volume = {25},
number = {7},
pages = {101604},
publisher = {American Society for Biochemistry and Molecular Biology},
issn = {1535-9476},
doi = {10.1016/j.mcpro.2026.101604},
url = {https://doi.org/10.1016/j.mcpro.2026.101604},
pmid = {42309337},
pmcid = {PMC13382583}
}

RIS

TY - JOUR
AU - Muñoz-Estrada, Jesús
AU - Mostafania, Andrew
AU - Halwatura, Lahiruni
AU - Haghani, Ali
AU - Jiang, Yuming
AU - Meyer, Jesse G
TI - Optimizing NGN2 Dosage Enhances the Neuronal Enrichment of iPSC-Derived Neuronal Cultures
T2 - Molecular & cellular proteomics : MCP
J2 - Mol Cell Proteomics
PY - 2026
DA - 2026/06/17
VL - 25
IS - 7
SP - 101604
SN - 1535-9476
PB - American Society for Biochemistry and Molecular Biology
DO - 10.1016/j.mcpro.2026.101604
UR - https://doi.org/10.1016/j.mcpro.2026.101604
LA - en
ER -

CSL-JSON

{
"id": "10.1016/j.mcpro.2026.101604",
"type": "article-journal",
"title": "Optimizing NGN2 Dosage Enhances the Neuronal Enrichment of iPSC-Derived Neuronal Cultures",
"container-title": "Molecular & cellular proteomics : MCP",
"author": [
{
"family": "Muñoz-Estrada",
"given": "Jesús"
},
{
"family": "Mostafania",
"given": "Andrew"
},
{
"family": "Halwatura",
"given": "Lahiruni"
},
{
"family": "Haghani",
"given": "Ali"
},
{
"family": "Jiang",
"given": "Yuming"
},
{
"family": "Meyer",
"given": "Jesse G"
}
],
"container-title-short": "Mol Cell Proteomics",
"volume": "25",
"issue": "7",
"page": "101604",
"DOI": "10.1016/j.mcpro.2026.101604",
"PMID": "42309337",
"PMCID": "PMC13382583",
"ISSN": "1535-9476",
"publisher": "American Society for Biochemistry and Molecular Biology",
"URL": "https://doi.org/10.1016/j.mcpro.2026.101604",
"language": "en",
"issued": {
"date-parts": [
[
2026,
6,
17
]
]
}
}

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:10.1093/nar/gkag368 [code]
Single-cell trajectory inference for detecting transient events in biological processes.
Journal: Nucleic acids research
In common: anndata, Scanpy, scikit-learn, 4 other tools, genetics / omics, cellular / molecular, 3 references, author Jesse G Meyer
[2] 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: anndata, Scanpy, Plotly, 6 other tools, genetics / omics, cellular / molecular, 2 references
[3] doi:10.1038/s41514-026-00391-9 [code]
Region-specific transcriptional signatures of brain aging in the absence of neuropathology at the single-cell level.
Journal: npj aging
In common: anndata, Scanpy, Plotly, 7 other tools, genetics / omics, cellular / molecular, 1 reference
[4] doi:10.1038/s41593-026-02267-3 [code]
Spatial proteomic analysis in human Alzheimer's disease brains enables identification of microenvironment-dependent microglial cell states.
Journal: Nature neuroscience
In common: anndata, Scanpy, Plotly, 7 other tools, genetics / omics, cellular / molecular
[5] doi:10.1126/sciadv.aeg3223 [code]
The extreme diversity of retinal amacrine cells has deep evolutionary roots.
Journal: Science advances
In common: anndata, Scanpy, Plotly, 6 other tools, genetics / omics, cellular / molecular, 1 reference
[6] doi:10.1371/journal.pcbi.1014346 [code]
StPedf: Cell trajectory inference of spatial transcriptomics via spatial proximity embedding and spatial density-adaptive fusion.
Journal: PLoS computational biology
In common: anndata, Scanpy, Plotly, 7 other tools, genetics / omics
[7] doi:10.1016/j.isci.2026.116055 [code]
Mapping the transcriptional diversity of calcium signaling in the mouse and human brain.
Journal: iScience
In common: anndata, Scanpy, Plotly, 7 other tools, genetics / omics
[8] doi:10.1016/j.celrep.2026.117270 [code]
Blocking apoptosis promotes survival and alters developmental dynamics of human retinal ganglion cells in retinal organoids.
Journal: Cell reports
In common: anndata, Scanpy, Plotly, 7 other tools, genetics / omics
[9] doi:10.3390/ijms27167275 [code]
Integrative Multi-Omics Analysis of Multiple Sclerosis Reveals Cell-Type-Specific Regulatory Landscapes and Discordant Methylation-Expression Coupling.
Journal: International journal of molecular sciences
In common: anndata, Scanpy, statsmodels, 6 other tools, genetics / omics, cellular / molecular, 1 reference
[10] doi:10.1038/s41467-026-76054-w [code]
Distinct differentiation trajectories leave lasting impacts on gene regulation and function of V2a neurons.
Journal: Nature communications
In common: Scanpy, Plotly, pandas, 2 other tools, cellular / molecular, 4 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.