OSCR

Multi-omics profiling-derived signature links cellular ecosystem to glioblastoma prognosis.

Code ↔ Paper

16 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 16 matches
  1. [1] § Results › Development and validation of an ECM-related prognostic signature in GBM ↔ P1_RNA_TCGA.r, lines 179–261 · score 0.97 · eukaryotic translation initiation, cell cycle checkpoints, cell projection membrane, neuron synapse, synaptic membrane, focal adhesion
  2. [2] § Results › Metabolic reprogramming in prognostically unfavorable Scissor-Positive cell populations ↔ P5_scRNA.r, lines 166–211 · score 0.76 · oxidative phosphorylation, scMetabolism, drug metabolism, gluconeogenesis, glycolysis, neg
  3. [3] § Results › Single-cell resolution of ECMSig expression and identification of prognostically relevant cellular states in GBM ↔ P5_scRNA.py, lines 60–125 · score 0.70 · CD3E, UMAP, CD68, NES, PECAM1, PTPRC
  4. [4] § STAR★Methods › Method details › Construction and validation of the prognostic signature ↔ P1_RNA_TCGA.r, lines 179–261 · score 0.68 · extracellular matrix, cell adhesion, positive regulation, collagen, TCGA, survival
  5. [5] § Results › Development and validation of an ECM-related prognostic signature in GBM ↔ P1_RNA_TCGA.r, lines 263–325 · score 0.63 · candidate genes, ECM related, high risk, GBM tumors, TCGA GBM, upregulated
  6. [6] § STAR★Methods › Method details › Single-cell RNA-seq analysis ↔ P5_scRNA.r, lines 166–211 · score 0.60 · scMetabolism, scRNA, Metabolic, phenotypic, Scissor, survival
  7. [7] § STAR★Methods › Method details › Drug sensitivity prediction ↔ R/CALCPHENOTYPE.R, lines 659–717 · score 0.57 · OncoPredict, Drug sensitivity, gene expression, trained, model, prediction
  8. [8] § Results › Single-cell resolution of ECMSig expression and identification of prognostically relevant cellular states in GBM ↔ P5_scRNA.py, lines 60–125 · score 0.55 · IL4I1, UMAP, AEBP1, CD81, glioma, pericytes
  9. [9] § Results › Transcriptomic and immune microenvironment features of ECMSig-stratified GBM ↔ P1_RNA_TCGA.r, lines 493–561 · score 0.54 · immune score xcell, Macrophage_XCELL, spearman, infiltration
  10. [10] § STAR★Methods › Method details › Construction and validation of the prognostic signature ↔ P4_Protein_CAPTC.r, lines 136–194 · score 0.53 · extracellular matrix structural, regulation, collagen, genes
  11. [11] § Results › Functional states and intercellular communication networks of prognostically detrimental cell subpopulations ↔ P5_scRNA.r, lines 46–103 · score 0.53 · CellChat, scissor pos, strength, interactions, myeloid, pathways
  12. [12] § STAR★Methods › Method details › Genomic alteration analysis ↔ R/IDWAS.R, lines 231–277 · score 0.52 · mutated genes, Somatic mutation, variants, TCGA, drug
  13. [13] § STAR★Methods › Method details › Genomic alteration analysis ↔ vignettes/glds.Rmd, lines 33–149 · score 0.52 · Somatic mutation, drug gene, variants, mutated, pathways
  14. [14] § Results › Development and validation of an ECM-related prognostic signature in GBM ↔ P5_scRNA.py, lines 1–57 · score 0.52 · IL4I1, CTSD, PCOLCE2, TMEM102, PLAUR, AEBP1
  15. [15] § Results › Development and validation of an ECM-related prognostic signature in GBM ↔ R/GLDS.R, lines 1–105 · score 0.51 · tuning parameter, cross validation, optimal, bars, LASSO, model
  16. [16] § Results › External validation of the prognostic value of ECMSig score ↔ P2_validate_survival.r, lines 53–59 · score 0.50 · CGGA_325, CGGA_693, clinical, survival, validation, cohort

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

R · 561 lines · 16 KB · CC-BY-4.0 · 4 matches

  1. library(tidyverse)
  2. library(glue)
  3. library(qs)
  4. library(ggplot2)
  5. library(patchwork)
  6. library(ggpubr)
  7. library(survival)
  8. library(survminer)
  9. library(clusterProfiler)
  10. library(org.Hs.eg.db)
  11. library(BSgenome.Hsapiens.UCSC.hg19)
  12. library(ReactomePA)
  13. library(viridis)
  14. library(glmnet)
  15. library(ggvenn)
  16. library(ComplexHeatmap)
  17. library(rms)
  18. library(edgeR)
  19. library(decoupleR)
  20. library(prettyunits)
  21. library(GseaVis)
  22. library(circlize)
  23. library(RColorBrewer)
  25. # tumor v.s. normal deg from GEPIA2
  27. gene_df = qread("data/Gene_info.qs")
  29. deg_df <- read.table("data/TCGA_GBM_GEPIA2_Tumor_Normal_deg.txt", header = T, sep = "\t") %>%
  30. rename_with(
  31. .fn = ~ c("Gene", "GeneID", "Tumor_median", "Normal_median", "Log2FC", "adjp"),
  32. .cols = 1:6
  33. ) %>%
  34. filter(Gene %in% gene_df$gene_name) %>%
  35. mutate(
  36. group = case_when(
  37. Log2FC >= 1 & adjp <= 0.05 ~ "Up",
  38. Log2FC <= -1 & adjp <= 0.05 ~ "Down",
  39. TRUE ~ "NotSignificant"
  40. )
  41. )
  43. # Down NotSignificant Up
  44. # 1771 10252 4367
  46. deg_df %>%
  47. ggplot(aes(x = Log2FC, y = -log(adjp), color = group)) +
  48. geom_point(alpha = 0.5) +
  49. labs(x = "log2(Tumor / Normal)", y = "-log(adjusted p-value)") +
  50. scale_color_manual(values = c("Down" = "#2d5ba1", "Up" = "#e43030", "NotSignificant" = "#9D9D9D")) +
  51. geom_vline(xintercept = 0, linetype = "dashed") +
  52. geom_hline(yintercept = 0, linetype = "dashed") +
  53. ggtitle("Tumor vs Normal Expression") +
  54. theme_bw() +
  55. theme(
  56. legend.position = "none",
  57. aspect.ratio = 1,
  58. plot.title = element_text(hjust = 0.5)
  59. )
  61. # Survival analysis of primary GBM
  63. final_surv_df = qread("data/TCGA_GBM_metainfo.qs")
  65. final_tpm = qread("data/TCGA_GBM_tpm.qs")
  67. final_surv_df <- cbind(
  68. final_surv_df,
  69. as.data.frame(t(final_tpm))
  70. )
  72. result_df <- data.frame()
  74. for (i in 7:ncol(final_surv_df)) {
  75. col_name <- colnames(final_surv_df)[i]
  76. tmp <- final_surv_df %>%
  77. mutate(
  78. group = factor(
  79. ifelse(get(col_name) > median(get(col_name)), "high", "low"),
  80. levels = c("low", "high")
  81. )
  82. )
  84. fit <- coxph(Surv(OS_time, OS_status) ~ group, data = tmp)
  86. hazard_fi <- log(summary(fit)$conf.int[, "exp(coef)"])
  87. p_fi <- summary(fit)$logtest["pvalue"]
  89. result_df <- rbind(
  90. result_df,
  91. data.frame(
  92. col_name = col_name,
  93. p_fi = p_fi,
  94. hazard_fi = hazard_fi
  95. )
  96. )
  97. }
  99. result_df %<>%
  100. mutate(hazard_fi = 10**hazard_fi) %>%
  101. filter(!is.na(hazard_fi)) %>%
  102. filter(p_fi < 0.05) %>%
  103. arrange(p_fi)
  105. filtered_genes <- colnames(filtered_surv_df)[-(1:6)][apply(filtered_surv_df[, -(1:6)], 2, median) != 0]
  107. result_df %<>%
  108. filter(col_name %in% filtered_genes) %>%
  109. rename_with(
  110. .fn = ~ c("Gene"),
  111. .cols = 1
  112. ) %>%
  113. mutate(
  114. group = case_when(
  115. hazard_fi > 1 & p_fi <= 0.05 ~ "poor",
  116. hazard_fi < 1 & p_fi <= 0.05 ~ "good",
  117. TRUE ~ "other"
  118. )
  119. ) %>%
  120. arrange(hazard_fi) %>%
  121. mutate(
  122. rank = row_number()
  123. )
  125. # good other poor
  126. # 194 16683 612
  128. ggplot(result_df, aes(x = rank, y = log(hazard_fi), fill = group, color = group)) +
  129. geom_point() +
  130. labs(x = "Rank", y = "log(hazard ratio)") +
  131. scale_color_manual(values = c("good" = "#2d5ba1", "poor" = "#e43030", "other" = "#9D9D9D")) +
  132. ggtitle("Tumor vs Normal Expression") +
  133. theme_bw() +
  134. theme(
  135. legend.position = "none",
  136. aspect.ratio = 1,
  137. plot.title = element_text(hjust = 0.5)
  138. )
  140. # enrichment analysis of survival genes and DEG genes
  142. result_df <- qread("res/TCGA-GBM_survival_gene.qs")
  143. deg_df <- qread("data/GBM_Tumor_Normal_GEPIA_Limma_deg.qs")
  145. deg_df$group <- deg_df$threshold
  147. get_enrich <- function(df, term = "NA", exclude_group = NULL) {
  149. odf <- data.frame()
  151. group <- setdiff(unique(df$group), exclude_group)
  152. for (i in group) {
  153. gs <- df[df$group == i, ]$Gene
  154. print(length(gs))
  156. trans_id <- bitr(gs, fromType = "SYMBOL", toType = c("ENTREZID", "ENSEMBL"), OrgDb = "org.Hs.eg.db")
  158. for (j in c("BP", "CC", "MF")) {
  159. go <- enrichGO(gs, OrgDb = org.Hs.eg.db, ont = j, keyType = "SYMBOL", pool = T, readable = F)
  160. go.res <- go@result
  161. go.res$path <- j
  162. go.res$group <- i
  163. go.res$term <- term
  165. odf <- rbind(odf, go.res)
  166. }
  168. react <- enrichPathway(gene = unique(trans_id$ENTREZID), readable = T)
  169. react.res <- react@result
  170. react.res$path <- "REACTOME"
  171. react.res$group <- i
  172. react.res$term <- term
  174. odf <- rbind(odf, react.res)
  175. }
  176. return(odf)
  177. }
  179. enrich_df <- rbind(
  180. get_enrich(result_df, term = "Survival", exclude_group = "other"),
  181. get_enrich(deg_df, term = "DEG", exclude_group = "NotSignificant")
  182. )
  184. term_list = list(
  186. 'Survival_poor' = c(
  187. 'collagen-containing extracellular matrix',
  188. 'basement membrane',
  189. 'axon terminus',
  190. 'perikaryon',
  191. 'neuron projection terminus',
  192. 'interstitial matrix',
  193. 'cell projection membrane'
  194. ),
  196. 'Survival_good' = c(
  197. 'eukaryotic translation initiation factor 3 complex',
  198. 'eukaryotic 48S preinitiation complex',
  199. 'eukaryotic 43S preinitiation complex',
  200. 'translation preinitiation complex',
  201. 'Formation of a pool of free 40S subunits',
  202. 'L13a-mediated translational silencing of Ceruloplasmin expression',
  203. 'GTP hydrolysis and joining of the 60S ribosomal subunit',
  204. 'Eukaryotic Translation Initiation',
  205. 'Cap-dependent Translation Initiation',
  206. 'formation of cytoplasmic translation initiation complex'
  207. ),
  209. 'DEG_Up' = c(
  210. 'focal adhesion',
  211. 'cell-substrate junction',
  212. 'Neutrophil degranulation',
  213. 'Interferon Signaling',
  214. 'collagen-containing extracellular matrix',
  215. 'Mitotic G1 phase and G1/S transition',
  216. 'Cell Cycle Checkpoints',
  217. 'regulation of innate immune response',
  218. 'leukocyte migration',
  219. 'positive regulation of cell adhesion'
  220. ),
  222. 'DEG_Down' = c(
  223. 'Neuronal System',
  224. 'synaptic membrane',
  225. 'neuron to neuron synapse',
  226. 'postsynaptic specialization',
  227. 'asymmetric synapse',
  228. 'postsynaptic density',
  229. 'postsynaptic membrane',
  230. 'Transmission across Chemical Synapses',
  231. 'regulation of membrane potential',
  232. 'monoatomic ion channel complex'
  233. )
  234. )
  236. term_df = data.frame(
  237. group = c("poor", "good", "Up", "Down"),
  238. term = c("Survival", "Survival", "DEG", "DEG")
  239. )
  241. for(i in 1:nrow(term_df)){
  242. s_group = term_df$group[i]
  243. s_term = term_df$term[i]
  245. sel_terms = term_list[[str_glue('{s_term}_{s_group}')]]
  247. enrich_df %>%
  248. filter(group == s_group & term == s_term) %>%
  249. filter(Description %in% sel_terms) %>%
  250. mutate(gene_ratio = as.numeric(GeneRatio %>% sapply(function(x) {parse(text = x) %>% eval() %>% round(3)}))) %>%
  251. arrange(- gene_ratio) %>%
  252. mutate(Description = factor(Description, levels = rev(Description))) %>%
  253. ggplot(aes(x = Description, y = gene_ratio, fill = -log10(qvalue), color = -log10(qvalue))) +
  254. geom_point(aes(size = Count))+
  255. scale_color_viridis(option = "inferno")+
  256. scale_fill_viridis(option = "inferno")+
  257. coord_flip()+
  258. theme_bw()+
  259. labs(y = NULL, x = 'Gene ratio')
  261. }
  263. # lasso model
  265. surv_gene_df <- qread("res/TCGA-GBM_survival_gene.qs")
  267. deg_df <- qread("data/GBM_Tumor_Normal_GEPIA_Limma_deg.qs")
  269. path_genes <- list.files("data/msigdb", full.names = T) %>%
  270. .[-6] %>%
  271. lapply(read.table, header = F, skip = 1) %>%
  272. do.call(rbind, .) %>%
  273. .[, 1] %>%
  274. as.character() %>%
  275. unique()
  277. genes_list <- list(
  278. 'High risk' = surv_gene_df %>% filter(group == "poor") %>% pull(Gene),
  279. 'Low risk' = surv_gene_df %>% filter(group == "good") %>% pull(Gene),
  280. 'Upregulated' = deg_df %>% filter(threshold == "Up") %>% pull(Gene),
  281. 'Downregulated' = deg_df %>% filter(threshold == "Down") %>% pull(Gene),
  282. 'ECM-related' = path_genes
  283. )
  285. ggvenn(genes_list[c(2,4,5)],
  286. fill_color = c("#2d5ba1", "#e43030", "#227d3c"),
  287. fill_alpha = 0.5,
  288. text_size = 3,
  289. set_name_size = 3,
  290. stroke_size = 0)
  292. candidate_genes <- Reduce(intersect,genes_list[c(1,3,5)])
  294. final_tpm_input <- qread("data/TCGA_GBM_tpm.qs") %>% t()
  296. final_surv_df <- qread("data/TCGA_GBM_meta.qs")
  298. rownames(final_surv_df) <- NULL
  300. final_surv_df_input <- final_surv_df[, 1:3] %>% column_to_rownames("sample")
  302. table(rownames(final_tpm_input) == rownames(final_surv_df_input))
  304. colnames(final_surv_df_input) <- c("time", "status")
  306. x <- as.matrix(final_tpm_input[, candidate_genes])
  307. y <- as.matrix(final_surv_df_input)
  309. fit <- glmnet(x, y, family = "cox")
  310. plot(fit, xvar = "lambda", label = FALSE)
  312. cvfit <- cv.glmnet(x, y, family = "cox", type.measure = "C")
  313. coef(cvfit, s = cvfit$lambda.1se)
  315. plot(cvfit, xvar = "lambda", label = FALSE)
  317. para_df <-
  318. coef(cvfit, s = cvfit$lambda.1se) %>%
  319. as.matrix() %>%
  320. as.data.frame() %>%
  321. rename_with(~ c('beta'), 1) %>%
  322. filter(beta != 0) %>%
  323. arrange( - beta)
  325. Heatmap(as.matrix(para_df),col = circlize::colorRamp2(c( 0, 0.03), c("white", "#b81313")),
  326. cell_fun = function(j, i, x, y, width, height, fill) {
  327. grid.text(sprintf("%.6f", as.matrix(para_df)[i, j]), x, y, gp = gpar(fontsize = 10))}
  328. )
  330. # multivariate
  332. lasso_genes <- para_df %>% rownames(.)
  334. model_expr <- para_df %>%
  335. mutate(expr = str_c(beta, "*", rownames(.))) %>%
  336. pull(expr) %>%
  337. paste(collapse = "+")
  339. multi_surv_df <- cbind(
  340. final_surv_df[,1:6],
  341. final_tpm_input[, lasso_genes]
  342. ) %>%
  343. mutate(
  344. score = eval(parse(text = model_expr), envir = .),
  345. score_group = factor(ifelse(score > median(score), "high", "low"), levels = c("low", "high"))
  346. )
  348. res.cox <- coxph(Surv(OS.time, OS) ~ score + gender + age, data = multi_surv_df)
  350. survminer::ggforest(res.cox,
  351. data = multi_surv_df,
  352. main = "Hazard ratio",
  353. fontsize = 1.0
  354. )
  356. # nomogram
  358. ddist <- datadist(multi_surv_df)
  359. options(datadist = "ddist")
  361. res.cox <- rms::cph(Surv(OS.time, OS) ~ score + age, data = multi_surv_df, x = T, y = T, surv = T)
  363. survival <- rms::Survival(res.cox)
  364. survival1 <- function(x) survival(365, x)
  365. survival2 <- function(x) survival(730, x)
  367. p <- rms::nomogram(res.cox,
  368. fun = list(survival1, survival2),
  369. funlabel = c("1 year survival", "2 year survival")
  370. )
  372. # score High v.s. Low
  374. tumor_count <- round(qread('data/TCGA_GBM_count.qs'))
  376. t.filter <- tumor_count[rowSums(tumor_count) >= 10, ]
  378. condi <- multi_surv_df[match(colnames(t.filter), multi_surv_df$sample), ]
  380. coldata <- data.frame(condition = factor(condi$score_group, levels = c('high','low')), row.names = condi$sample)
  382. group_list <- coldata$condition
  384. design <- model.matrix(~0+group_list)
  386. rownames(design) <- colnames(t.filter)
  387. colnames(design) <- levels(coldata$condition)
  389. deglist <- edgeR::DGEList(t.filter, group = coldata$condition)
  391. deglist <- calcNormFactors(deglist)
  393. contrast.matrix<-makeContrasts(paste0(unique(group_list),collapse = "-"),levels = design)
  395. fit <- lmFit(t.filter,design)
  396. fit2 <- contrasts.fit(fit, contrast.matrix)
  397. fit2 <- eBayes(fit2)
  398. tempOutput = topTable(fit2, coef=1, n=Inf)
  399. nrDEG = na.omit(tempOutput)
  401. nrDEG <- nrDEG %>%
  402. mutate(threshold = case_when(
  403. adj.P.Val < 0.05 & 2**logFC >= 1.2 ~ "Up",
  404. adj.P.Val < 0.05 & 2**logFC <= 1/1.2 ~ "Down",
  405. TRUE ~ "NotSignificant"
  406. )) %>%
  407. arrange( - logFC)
  409. ggplot(nrDEG, aes(x = logFC, y = -log(adj.P.Val), color = threshold)) +
  410. geom_point(alpha = 0.5) +
  411. labs(x = "log2(High / Low)", y = "-log(adjusted p-value)") +
  412. scale_color_manual(values = c("Down" = "#2d5ba1", "Up" = "#e43030", "NotSignificant" = "#9D9D9DFF")) +
  413. geom_vline(xintercept = 0, linetype = "dashed") +
  414. geom_hline(yintercept = 0, linetype = "dashed") +
  415. ggtitle("High vs Low Expression") +
  416. theme_bw() +
  417. theme(
  418. legend.position = "none",
  419. aspect.ratio = 1,
  420. plot.title = element_text(hjust = 0.5)
  421. )
  423. net <- decoupleR::get_progeny(organism = 'human', top = 500)
  425. sample_acts <- decoupleR::run_mlm(mat = final_tpm_input, net = net, .source = 'source', .target = 'target',.mor = 'weight', minsize = 5)
  427. sample_acts_mat <- sample_acts %>%
  428. tidyr::pivot_wider(id_cols = 'condition',
  429. names_from = 'source',
  430. values_from = 'score') %>%
  431. tibble::column_to_rownames('condition') %>%
  432. as.matrix()
  434. sample_acts_mat <- scale(sample_acts_mat)
  436. colors <- rev(RColorBrewer::brewer.pal(n = 11, name = "RdBu"))
  437. colors.use <- grDevices::colorRampPalette(colors = colors)(100)
  439. my_breaks <- c(seq(-3, 0, length.out = ceiling(100 / 2) + 1),
  440. seq(0.05,3, length.out = floor(100 / 2)))
  442. multi_surv_df <- qread( "data/TCGA_GBM_meta.qs")
  444. anno_df <- multi_surv_df[match(rownames(sample_acts_mat),multi_surv_df$sample),c('sample','score','score_group')] %>% arrange(score)
  445. rownames(anno_df) <- anno_df$sample
  446. anno_df <- anno_df[,2:3]
  448. p=pheatmap::pheatmap(mat = t(sample_acts_mat)[,rownames(anno_df)],
  449. annotation_col = anno_df,
  450. cluster_cols = F,
  451. annotation_colors = list(
  452. score_group = c('high' = '#D02324', low = '#3161A7'),
  453. score = c( "white", "#FD8235")
  454. ),
  455. show_colnames = F,
  456. breaks = my_breaks,
  458. color = colors.use,
  459. border_color = "white")
  461. hallmark_geneset <- read.gmt('data/h.all.v2024.1.Hs.symbols.gmt')
  463. nrDEG <- qread('res/score_H_L.deg.qs') %>% arrange( - logFC)
  465. geneList <- nrDEG$logFC
  466. names(geneList) <- rownames(nrDEG)
  468. GSEA_enrichment <- GSEA(geneList,
  469. TERM2GENE = hallmark_geneset,
  470. pvalueCutoff = 0.05,
  471. minGSSize = 10,
  472. maxGSSize = 500,
  473. eps = 0,
  474. pAdjustMethod = "BH")
  476. result <- data.frame(GSEA_enrichment)
  478. gseaNb(object = GSEA_enrichment,
  479. curveCol = c('#AADEC6','#FDC6A2','#C4CFE5','#F2C3E0','#FFF0A4','#4095CE','#8D60E0'),
  480. geneSetID = result$ID[c(2:5,14,13,15)])
  482. immune_df <- read.table('data/TIMER2_infiltration_estimation_for_tcga.csv.gz',sep=',',header=T) %>%
  483. dplyr::rename(sample = 'cell_type') %>%
  484. mutate(sample = str_replace_all(sample, '-', '.') %>% paste0('A')) %>%
  485. filter(sample %in% multi_surv_df$sample) %>%
  486. mutate(
  487. score = multi_surv_df[match(sample, multi_surv_df$sample),]$score,
  488. score_group = multi_surv_df[match(sample, multi_surv_df$sample),]$score_group,
  489. ) %>%
  490. column_to_rownames(var = 'sample') %>%
  491. arrange( score)
  493. immune_df_sub <- immune_df[,grepl('XCELL$',colnames(immune_df))] %>% scale() %>% t()
  495. rownames(immune_df_sub) <- str_remove_all(rownames(immune_df_sub), '_XCELL')
  497. rownames(immune_df_sub) %<>% str_replace_all(., '\\.+', ' ')
  499. immune_df_sub <- immune_df_sub[apply(immune_df_sub > 0, 1, sum) > 14,]
  501. for(i in rownames(immune_df_sub)){
  502. low_score = immune_df_sub[i,rownames(immune_df[immune_df$score_group == 'low',])] %>% as.numeric()
  503. high_score = immune_df_sub[i,rownames(immune_df[immune_df$score_group == 'high',])] %>% as.numeric()
  505. score_test = wilcox.test(high_score,low_score)
  507. if(score_test$p.value < 0.05){
  508. x = case_when(
  509. score_test$p.value > 0.01 ~ '*',
  510. score_test$p.value > 0.001 ~ '**',
  511. TRUE ~ '***'
  512. )
  514. if(median(high_score) > median(low_score)){
  515. print(str_glue("{i}: up in high score, {x}"))
  516. }else{
  517. print(str_glue("{i}: up in low score, {x}"))
  518. }
  519. }
  520. }
  523. Heatmap(immune_df_sub,
  524. top_annotation = HeatmapAnnotation(
  525. score = immune_df$score,
  526. score_group = immune_df$score_group,
  527. col = list(
  528. score_group = c('high' = '#D02324', low = '#3161A7'),
  529. score = colorRamp2(c(0.5, 3), c( "white", "#FD8235"))
  530. )
  531. ),
  532. name = "Immune infiltrates",
  533. col = colorRamp2(c(-1, 0, 1), c("blue", "white", "red")),
  534. clustering_distance_rows = "spearman",
  535. clustering_method_rows = 'ward.D2',
  536. cluster_rows = T,
  537. cluster_columns = F,
  538. show_column_names = F,
  539. row_names_gp = gpar(fontsize = 10),
  540. column_names_gp = gpar(fontsize = 10),
  541. show_row_names = T)
  544. immune_df %>%
  545. dplyr::select('immune.score_XCELL','Macrophage_XCELL','score') %>%
  546. gather(key = 'cell', value = 'infiltration', - score) %>%
  547. ggscatter(x = "score", y = "infiltration",
  548. shape=21,
  549. color = 'black',
  550. fill="transparent",
  551. size = 3,
  552. add = "reg.line",
  553. add.params = list(color = "skyblue4", fill = "lightskyblue1"),
  554. conf.int = TRUE,
  555. cor.coef = TRUE,
  556. cor.coeff.args = list(method = "spearman",
  557. label.x = 1.0, label.y= 0.1,
  558. label.sep = "\n")
  559. )+
  560. facet_wrap(~cell, ncol = 2, scales = "free_y")+
  561. theme(aspect.ratio = 1)

P1_RNA_TCGA.r, under CC-BY-4.0 · at the source

Overview

Authors: Zhen Zhang1, Hao Xu2, Haijing Zheng3, Zhaolong Pan3, Mei Feng4, Yongchang Yang5, Manqing Cao2
  1. Department of Neuro-Oncology and Neurosurgery, Tianjin Medical University Cancer Institute & Hospital, National Clinical Research Center for Cancer, Key Laboratory of Cancer Prevention and Therapy, Tianjin’s Clinical Research Center for Cancer, Tianjin 300060, China
  2. The Second Surgical Department of Breast Cancer, Tianjin Medical University Cancer Institute & Hospital, National Clinical Research Center for Cancer, Tianjin’s Clinical Research Center for Cancer, Key Laboratory of Breast Cancer Prevention and Therapy, Tianjin Medical University, Ministry of Education, Key Laboratory of Cancer Prevention and Therapy, Tianjin 300060, China
  3. Department of Hepatobiliary Cancer, Research Center for Prevention and Treatment of Liver Cancer, Tianjin Medical University Cancer Institute & Hospital, National Clinical Research Center for Cancer, Tianjin’s Clinical Research Center for Cancer, Tianjin Key Laboratory of Digestive Cancer, Tianjin, China
  4. Division of General Surgery, Peking University First Hospital, Peking University, No. 8 Xi Shiku Street, Beijing 100034, China
  5. Department of Radiation Oncology, The First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi 530022, China
Institutions: Tianjin Medical University Cancer Institute and Hospital (China); Tianjin Medical University (China); Peking University (China); Peking University First Hospital (China); Guangxi Medical University (China); First Affiliated Hospital of GuangXi Medical University (China)
Journal: iScience, volume 29, issue 6, article 115982
Dates: received 19 August 2025; accepted 29 April 2026; published online 18 May 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1016/j.isci.2026.115982 · PMID 42181228 · PMCID PMC13197639 · OpenAlex W7161537316
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: genetics / omics (modality), human (organism), other condition (population)
Methods: Statistics, Connectivity
Keywords: Disease, Oncology, Transcriptomics
Topic: Glioma Diagnosis and Treatment (Genetics, Medicine), according to OpenAlex
Funding: National Postdoctoral Program for Innovative Talents (BX20240026); National Natural Science Foundation of China (82403722); China Postdoctoral Science Foundation (2024M762383)
Citations: not cited yet (Europe PMC); 45 references in the paper
Research resources: KNS42 RRID:CVCL_0378, KALS1 RRID:CVCL_1323, KS1 RRID:CVCL_1343, LN-428 RRID:CVCL_3959

Abstract

Glioblastoma (GBM) remains a devastating brain malignancy with a dismal prognosis, underscoring the urgent need for robust prognostic biomarkers and therapeutic targets. Here, we developed and validated a seven-gene extracellular matrix-related prognostic signature (ECMSig) using multi-omics data. The ECMSig robustly stratified GBM patients into high- and low-risk groups with distinct overall survival in The Cancer Genome Atlas cohort and Chinese Glioma Genome Atlas cohorts. High ECMSig scores were associated with aggressive molecular features, including upregulation of epithelial-mesenchymal transition and hypoxia, and a tumor-promoting immune microenvironment. Single-cell RNA sequencing analysis identified prognostic Scissor-Positive tumor, myeloid, and endothelial cells exhibiting high ECMSig scores, mesenchymal/immunosuppressive phenotypes, and notable metabolic reprogramming. These cells orchestrate a complex intercellular communication network and spatially co-localize within hypoxic perivascular niches. Furthermore, ECMSig predicted differential drug sensitivities, offering potential therapeutic avenues. The prognostic ECMSig highlights the complex interplay within the GBM ecosystem, paving the way for personalized therapeutic strategies.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repositories

Its files are read in the Code ↔ Paper reader above, with 16 matches between paragraphs and lines of code.

Zenodo 17669213

License: CC-BY-4.0
State: the link answers, verified on 28 September 2026
Evidence: files inventoried
Languages: R (7), Python (2)
Size: 9 files, 9 scripts
Software Heritage: not checked
Found in: “Data and code availability”
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: tidyverse (7 files), ggpubr (4 files), survival (4 files), circlize (2 files), clusterProfiler (2 files), ComplexHeatmap (2 files), ggplot2 (2 files), Matplotlib (2 files), patchwork (2 files), pheatmap (2 files), Scanpy (2 files), edgeR (1 file), glmnet (1 file), limma (1 file), NumPy (1 file), pandas (1 file), Seurat (1 file)
Availability: 1 check, the latest on 28 September 2026: the link answers (HTTP 200)
  • 28 September 2026: the link answers (HTTP 200)
9 files
At the source: doi.org/10.5281/zenodo.17669213

HuangLabUMN/oncoPredict

License: none: the authors keep all their rights
State: the link answers, verified on 28 September 2026
Evidence: files inventoried
Commit: c16c1ffaffdd855f4ada4c83a3948e5c16b618c5, 13 June 2026
Languages: R (11)
Size: 36 files, 11 scripts
Software Heritage: archived
Found in: the resources table
Holds: README, environment (DESCRIPTION), tests, documentation, 4 notebooks
Not found: license file, CITATION.cff, continuous integration
Tools: car (1 file), glmnet (1 file), limma (1 file)
Availability: 1 check, the latest on 28 September 2026: the link answers
  • 28 September 2026: the link answers
12 files
At the source: github.com/HuangLabUMN/oncoPredict

The paper's code and data availability statement is in the Data section.

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:

  • 2 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 20 scripts, each with its path and the digest of its content;
  • 16 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

Datasets cited

Data and code availability

The bulk RNA sequencing (RNA-seq) data and matched clinical information can be accessed through the UCSC Xena website32 (GDC TCGA-GBM cohort, TPM) and the Chinese Glioma Genome Atlas (mRNA_693 and mRNA_325 cohorts).33,34 The genomic data of TCGA GBM cohort was available at UCSC Xena website. The GBM proteomic dataset and paired clinical information were accessed at the Clinical Proteomic Tumor Analysis Consortium (CPTAC).35 The single-cell transcriptomic sequencing dataset utilizing technology from the 10X Genomics platform was available under the accession number GEO: GSE182109 (https://ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE182109) at the Gene Expression Omnibus (GEO) repository.36 The spatial transcriptomic sequencing dataset using the 10X Genomics Visium platform was under accession GEO: GSE194329 (https://ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE194329).37 The RNA expression data of GBM cell lines were downloaded from DEPMAP website.20

All analysis scripts, custom functions, and visualization code publicly available at zenodo: https://doi.org/10.5281/zenodo.17669213.

Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request.

Reproduced under the paper's license (CC BY), from the paper cited above.

Materials availability

There are no additional data, software, databases, or applications/tools available beyond those disclosed in the current study. All data are included in the article and supplementary data section.

Reproduced under the paper's license (CC BY), from the paper cited above.

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, 28 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 7 authors, 3 keywords, 3 funders, 45 references, 4 RRIDs.

Cite

This paper

Zhang, Z., Xu, H., Zheng, H., Pan, Z., Feng, M., Yang, Y., & Cao, M. (2026). Multi-omics profiling-derived signature links cellular ecosystem to glioblastoma prognosis. iScience, 29(6), 115982. https://doi.org/10.1016/j.isci.2026.115982

BibTeX

@article{zhang2026multi,
author = {Zhang, Zhen and Xu, Hao and Zheng, Haijing and Pan, Zhaolong and Feng, Mei and Yang, Yongchang and Cao, Manqing},
title = {{Multi-omics profiling-derived signature links cellular ecosystem to glioblastoma prognosis}},
journal = {iScience},
year = {2026},
month = may,
volume = {29},
number = {6},
pages = {115982},
publisher = {Elsevier},
issn = {2589-0042},
doi = {10.1016/j.isci.2026.115982},
url = {https://doi.org/10.1016/j.isci.2026.115982},
pmid = {42181228},
pmcid = {PMC13197639}
}

RIS

TY - JOUR
AU - Zhang, Zhen
AU - Xu, Hao
AU - Zheng, Haijing
AU - Pan, Zhaolong
AU - Feng, Mei
AU - Yang, Yongchang
AU - Cao, Manqing
TI - Multi-omics profiling-derived signature links cellular ecosystem to glioblastoma prognosis
T2 - iScience
J2 - iScience
PY - 2026
DA - 2026/05/18
VL - 29
IS - 6
SP - 115982
SN - 2589-0042
PB - Elsevier
DO - 10.1016/j.isci.2026.115982
UR - https://doi.org/10.1016/j.isci.2026.115982
LA - en
ER -

CSL-JSON

{
"id": "10.1016/j.isci.2026.115982",
"type": "article-journal",
"title": "Multi-omics profiling-derived signature links cellular ecosystem to glioblastoma prognosis",
"container-title": "iScience",
"author": [
{
"family": "Zhang",
"given": "Zhen"
},
{
"family": "Xu",
"given": "Hao"
},
{
"family": "Zheng",
"given": "Haijing"
},
{
"family": "Pan",
"given": "Zhaolong"
},
{
"family": "Feng",
"given": "Mei"
},
{
"family": "Yang",
"given": "Yongchang"
},
{
"family": "Cao",
"given": "Manqing"
}
],
"container-title-short": "iScience",
"volume": "29",
"issue": "6",
"page": "115982",
"DOI": "10.1016/j.isci.2026.115982",
"PMID": "42181228",
"PMCID": "PMC13197639",
"ISSN": "2589-0042",
"publisher": "Elsevier",
"URL": "https://doi.org/10.1016/j.isci.2026.115982",
"language": "en",
"issued": {
"date-parts": [
[
2026,
5,
18
]
]
}
}

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: edgeR, limma, car, 13 other tools, genetics / omics, other condition, 4 references
[2] doi:10.1016/j.isci.2026.115657 [code]
Integration of machine learning to develop a disulfidptosis model for predicting glioma prognosis, immunotherapy response, and drug.
Journal: iScience
In common: glmnet, survival, edgeR, 12 other tools, other condition, 1 reference
[3] doi:10.1016/j.xcrm.2026.102682 [code]
TET CpG sequence-context-specific DNA demethylation shapes progression of IDH-mutant gliomas.
Journal: Cell reports. Medicine
In common: glmnet, survival, edgeR, 12 other tools, genetics / omics, other condition
[4] doi:10.1038/s42003-026-10957-8 [code]
Brain defence by the extracellular matrix protein Cochlin.
Journal: Communications biology
In common: glmnet, survival, edgeR, 13 other tools
[5] doi:10.1002/imt2.70163 [code]
Spatial multi-omics unveils sphingolipid metabolic reprogramming within the retinal pathological niche.
Journal: iMeta
In common: edgeR, limma, circlize, 12 other tools, genetics / omics, 2 references
[6] doi:10.1038/s41586-026-10214-2 [code]
Multidimensional profiling of heterogeneity in supratentorial ependymomas.
Journal: Nature
In common: edgeR, car, circlize, 12 other tools, genetics / omics, other condition, 1 reference
[7] doi:10.1093/bioinformatics/btag592 [code]
Network-based stratification of allele-specific expression reveals patient subgroups in Huntington's disease.
Journal: Bioinformatics (Oxford, England)
In common: survival, edgeR, limma, 12 other tools, genetics / omics, other condition
[8] doi:10.1016/j.cell.2026.05.026 [code]
The critical role of the endogenous immune compartment after CAR T cell therapy in recurrent GBM.
Journal: Cell
In common: survival, edgeR, limma, 10 other tools, genetics / omics, other condition, 2 references
[9] 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: edgeR, limma, circlize, 12 other tools
[10] doi:10.3390/ijms27104466 [code]
Uncovering the Key Circuit FOSL2/FOS/EGR3/EGR1, Contributing to the Hyperexcitability of Excitatory Neurons in the Epileptic Temporal Cortex and Hippocampus.
Journal: International journal of molecular sciences
In common: glmnet, limma, circlize, 10 other tools, genetics / omics

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.