OSCR

HNRNPH1 drives glioblastoma progression by regulating the splicing of cell cycle genes.

Code ↔ Paper

2 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 2 matches
  1. [1] § Materials and Methods › Analysis of spatial transcriptomics data ↔ Module1_MP.R, lines 56 to 99 · score 0.91 · FindVariableFeatures, NormalizeData, ScaleData, RunPCA, Seurat, assignment · Report a mismatch
  2. [2] § Materials and Methods › Analysis of single-cell RNA sequencing data ↔ Module1_MP.R, lines 56 to 99 · score 0.76 · FindClusters, FindNeighbors, RunPCA, Seurat, variable, log · Report a mismatch

Paper

Loaded from Europe PMC by your browser; OSCR does not store it: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

R · 650 lines · 32 KB · no license · 2 matches

  2. Sys.setenv(RETICULATE_PYTHON = "/.conda/envs/leiden/bin/python")
  3. Sys.setenv(RETICULATE_PYTHON = "/.conda/envs/leiden/bin/python")
  5. library(reticulate)
  6. library(Seurat)
  7. library(scalop)
  8. library(leiden)
  9. library(ggplot2)
  10. library(RColorBrewer)
  11. library(viridis)
  12. library(scales)
  13. library(reshape2)
  14. library(reshape2)
  15. library(scales)
  16. library(NMF)
  17. library(MetBrewer)
  18. library(colorspace)
  19. library(tibble)
  20. library(dplyr)
  21. library(data.table)
  22. library(stringr)
  23. library(readr)
  24. library(Matrix)
  25. library(bigmemory)
  26. library(doMC)
  27. library(patchwork)
  29. ###### Per sample Leiden clustering##############################
  31. samples_names <- (read.delim("general/GBM_samples.txt", header = FALSE, sep = "\t"))$V1
  33. # set parameters
  34. complexity_filter <- 1000
  35. mitochondrial_filter <- 20
  36. genes_filter <- 7000
  37. n_dim <- 20
  38. dim_filter <- 5^(-15)
  39. res_param <- 1
  40. sig_th <- .005
  41. mp_num <- 13
  42. distinct16_pal<-c("#11A579","#F2B701","#66C5CC","#80BA5A","#F6CF71","#7F3C8D","#CF1C90","#3969AC","#f97b72","#E73F74","#4b4b8f","#ACA4E2","#31C53F","#B95FBB","#D4D915","#28CECA")
  43. dim2use_list <- c(8,20,19,9,12,14,16,10,10,14,8,10,12,16,7,5,14,16,4,12,6,6,8,12,13,10) # set by jackstraw
  45. leiden_clustering <- sapply(c(1:length(samples_names)), function(i){
  46. print(samples_names[i])
  47. # load spatial data
  48. unfilt_obj<-Load10X_Spatial(data.dir = paste("general/GBM_data/",samples_names[i],"/outs", sep = ""),
  49. filename = "filtered_feature_bc_matrix.h5",
  50. assay = "Spatial",
  51. slice = "detected_tissue_image.jpg",
  52. filter.matrix = TRUE,
  53. to.upper = FALSE
  54. )
  56. # filtering
  57. exp_obj <- subset(unfilt_obj, subset = nCount_Spatial > complexity_filter) # filter out spots with less than #complexity_filter UMIs
  58. exp_obj[["percent.mt"]] <- PercentageFeatureSet(exp_obj, pattern = "^MT-") # filter spots with high percent (mitochondrial_filter) mitochondrial genes
  59. exp_obj <- subset(exp_obj, subset = percent.mt<mitochondrial_filter)
  61. # normalization and centering
  62. exp_obj <- NormalizeData(exp_obj, assay = "Spatial", normalization.method="LogNormalize", scale.factor=10000)
  63. exp_obj <- FindVariableFeatures(exp_obj, selection.method = "vst", nfeatures = genes_filter) # use only #genes_filter most variable genes
  64. exp_obj <- ScaleData(exp_obj)
  66. # PCA
  67. exp_obj <- RunPCA(exp_obj, features = VariableFeatures(object = exp_obj))
  68. #PCA dims to use determined by Jackstraw
  69. #exp_obj <- JackStraw(exp_obj, num.replicate = 100)
  70. #exp_obj <- ScoreJackStraw(exp_obj, dims = 1:n_dim)
  71. #js_scores <- exp_obj@reductions$pca@jackstraw$overall.p.values
  72. #dim2use <- max(js_scores[,"PC"][js_scores[,"Score"] < dim_filter])
  73. dim2use <- dim2use_list[i]
  75. # clustering
  76. exp_obj <- FindNeighbors(exp_obj, dims = 1:dim2use)
  77. exp_obj <- FindClusters(exp_obj, algorithm=4, resolution = res_param) # leiden based clustering
  79. spatial_sample_programs <- FindAllMarkers(exp_obj, only.pos = TRUE, return.thresh = sig_th)
  81. genes_num <- table(spatial_sample_programs$cluster)
  82. spots_clusters <- [email hidden][,"seurat_clusters"]
  83. names(spots_clusters) <- row.names([email hidden])
  84. cluster_num <- table([email hidden]$seurat_clusters)
  86. return(list(spatial_sample_programs,genes_num,spots_clusters,cluster_num)) # output: #genes per cluster, #spots per cluster, clusters genes sig, spots assignment to cluster
  88. })
  91. # clusters quality
  92. genes_num <- c()
  93. spots_num <- c()
  94. sapply(c(1:length(samples_names)), function(i){
  95. genes_num <<- c(genes_num,as.numeric(leiden_clustering[2,i][[1]]))
  96. spots_num <<- c(spots_num,as.numeric(leiden_clustering[4,i][[1]]))
  97. })
  98. hist(genes_num, breaks = 30)
  99. hist(spots_num, breaks = 30)
  101. # create genes sig table for save
  102. genes_sig <- leiden_clustering[1,1][[1]]
  103. genes_sig$sample <- rep(samples_names[1],dim(genes_sig)[1])
  105. sapply(c(2:length(samples_names)),function(i){
  106. genes_sig_temp <- leiden_clustering[1,i][[1]]
  107. genes_sig_temp$sample <- rep(samples_names[i],dim(genes_sig_temp)[1])
  108. genes_sig <<- rbind(genes_sig,genes_sig_temp)
  109. })
  111. # create clusters assignment table for save
  112. spots_assign <- as.data.frame(leiden_clustering[3,1][[1]])
  113. spots_assign$sample <- rep(samples_names[1],dim(spots_assign)[1])
  114. row.names(spots_assign) <- paste(row.names(spots_assign), "_", samples_names[1], sep = "")
  115. colnames(spots_assign) <- c("cluster","sample")
  117. sapply(c(2:length(samples_names)),function(i){
  118. spots_assign_temp <- as.data.frame(leiden_clustering[3,i][[1]])
  119. spots_assign_temp$sample <- rep(samples_names[i],dim(spots_assign_temp)[1])
  120. row.names(spots_assign_temp) <- paste(row.names(spots_assign_temp), "_", samples_names[i], sep = "")
  121. colnames(spots_assign_temp) <- c("cluster","sample")
  122. spots_assign <<- rbind(spots_assign,spots_assign_temp)
  123. })
  125. # set gene signatures list in the right format
  126. genes_list <- lapply(c(1:length(samples_names)), function(i){
  127. sample_table <- leiden_clustering[1,i][[1]]
  128. sample_genes <- lapply(c(1:length(unique(sample_table$cluster))),function(j){
  129. cluster_table <- sample_table[sample_table$cluster == j,]
  130. genes <- cluster_table$gene[order(cluster_table$avg_log2FC, decreasing = TRUE)][1:min(50,nrow(cluster_table))]
  131. return(genes)
  132. })
  133. return(sample_genes)
  134. })
  135. genes_list <- unlist(genes_list, recursive = FALSE)
  136. names(genes_list) <- unlist(sapply(c(1:length(samples_names)),function(i){
  137. paste(samples_names[i],c(1:length(leiden_clustering[4,i][[1]])), sep = "_")
  138. }))
  140. samples_list <- unlist(sapply(c(1:length(samples_names)),function(i){
  141. rep(samples_names[i],length(leiden_clustering[4,i][[1]]))
  142. }))
  143. #saveRDS(genes_list,"/home/labs/tirosh/spatial_glioma/results/leiden_MPs/all_leiden_programs_v2.rds")
  144. # generate an ordered jaccard matrix and clusters
  145. jac_mat = Jaccard(genes_list)
  146. jac_matord = scalop::hca_reorder(jac_mat)
  147. jac_plot <- melt(jac_matord)
  149. prog_ordered <- colnames(jac_matord)
  150. clusters <- scalop::hca_groups(jac_matord,
  151. cor.method="none",
  152. k=mp_num,
  153. min.size=4,
  154. max.size=0.5)
  156. clusterCut <- sapply(names(clusters), function(i){
  157. new_vec <- rep(i,length(clusters[[i]]))
  158. names(new_vec) <- clusters[[i]]
  159. return(new_vec)
  160. })
  162. names(clusterCut) <- NULL
  163. clusterCut <- unlist(clusterCut)
  165. # plot jaccard
  167. jac_plot$value[jac_plot$value >= 0.35] <- 0.35
  168. p1 <- ggplot(data = jac_plot, aes(x = Var1, y = Var2, fill = value)) + geom_raster() +
  169. scale_fill_gradientn(colours = colorspace::sequential_hcl(9, palette="YlOrRd", rev=T), limits = c(0,0.35), name = "Jaccard\nIndex") +
  170. theme(axis.ticks = element_blank(), panel.background = element_rect(fill = "white"), axis.line = element_blank(), legend.text.align = 0.5) +
  171. scale_x_discrete(name = "\nPrograms", labels = element_blank(), breaks = element_blank()) + scale_y_discrete(name = "\nPrograms", labels = element_blank(), breaks = element_blank())
  174. metaprog_df <- data.frame(row.names = prog_ordered,
  175. cluster = as.factor(clusterCut[prog_ordered]),
  176. samples = stringr::str_replace(prog_ordered, "_.*",""),
  177. cells = prog_ordered)
  179. samples_annotation_plot <- ggplot(metaprog_df, aes(x = factor(cells, levels = cells), y = 1)) +
  180. geom_raster(aes(fill = samples)) + theme(legend.position = "none",axis.ticks = element_blank(), panel.background = element_rect(fill = "white"), axis.text = element_blank(), axis.line = element_blank(), axis.text.x = element_blank()) +
  181. labs(x = "", y = "samples", fill = "")
  182. metaprog_annotation_plot <- ggplot(metaprog_df, aes(x = factor(cells, levels = cells), y = 1)) +
  183. geom_raster(aes(fill = cluster)) + theme(legend.position = "none", axis.ticks = element_blank(), panel.background = element_rect(fill = "white"), axis.text = element_blank(), axis.line = element_blank(), axis.text.x = element_blank()) +
  184. labs(x = "", y = "", fill = "")
  185. MP_plot <- egg::ggarrange(p1, metaprog_annotation_plot, samples_annotation_plot, ncol = 1, nrow = 3, heights = c(40, 2, 1))
  188. #Define meta-programs from program clusters, Sort by frequency across programs in cluster (>= 3)
  189. cut_off <- table(metaprog_df$cluster)*0.25
  190. cut_off[cut_off < 3] <- 3
  192. mp_freqs = sapply(clusters, function(k) sort(table(unlist(genes_list[k])),decreasing = T),simplify = F)
  193. metaprograms = sapply(names(mp_freqs), function(c) head(names(mp_freqs[[c]])[mp_freqs[[c]] >= as.numeric(cut_off[c])], 50), simplify = F)
  194. names(metaprograms)<-c("vascular","neuron","hypoxia","hypoxia.immune","oligo","AC.immune","low.quality","NPC2","macrophage","MES.stromal","OPC.AC","metabolism","MES1.RGL")
  197. metaprog_df$samp_cluster <- str_split(metaprog_df$cells, "_", simplify = T)[,2]
  198. metaprog_df$anno_cluster <- metaprog_df$cluster
  199. class(metaprog_df$anno_cluster)
  200. levels(metaprog_df$anno_cluster) <- c("vascular","MES.stromal","OPC.AC","metabolism","MES1.RGL","neuron","hypoxia","hypoxia.immune","oligo","AC.immune","low.quality","NPC2","macrophage")
  202. sapply(c(1:26), function(i){
  203. print(samples_names[i])
  204. spots_gen_assign <- as.data.frame(leiden_clustering[3,i][[1]])
  205. spots_gen_assign$generalized <- apply(spots_gen_assign,1,function(r){
  206. return(metaprog_df$anno_cluster[metaprog_df$samples == samples_names[i] & metaprog_df$samp_cluster == r])
  207. })
  208. spots_gen_assign$barcodes <- row.names(spots_gen_assign)
  209. spots_gen_assign <- spots_gen_assign[,c("barcodes", "generalized")]
  210. })
  213. ######generating input matrices for NMF
  214. # vector of file paths
  215. file_paths <- list.files(path ="/general/exp_mats_GBM/", pattern = "\\.rds", full.names = TRUE)#directory with counts mats
  216. sample_ls <- gsub(pattern = "\\.rds$", replacement = "", x = basename(file_paths))
  217. per_sample_mat <- lapply(file_paths, readRDS)
  219. for (i in seq_along(per_sample_mat)){
  220. m <- as.matrix(per_sample_mat[[i]])
  221. if(min(colSums(m)) == 0){m <- m[, colSums(m) != 0]}
  222. scaling_factor <- 1000000/colSums(m)
  223. m_CPM <- sweep(m, MARGIN = 2, STATS = scaling_factor, FUN = "*")
  224. m_loged <- log2(1 + (m_CPM/10))
  226. # removing genes with zero variance across all cells
  227. var_filter <- apply(m_loged, 1, var)
  228. m_proc <- m_loged[var_filter != 0, ]
  229. # filtering out lowly expressed genes
  230. exp_genes <- rownames(m_proc)[(rowMeans(m_proc) > 0.4)]
  231. m_proc <- m_proc[exp_genes, ]
  233. # centering data gene-wise
  234. count_mat_cent<- m_proc - rowMeans(m_proc)
  235. #nmf preprocessing
  236. count_mat_nmf<- count_mat_cent
  237. count_mat_nmf[count_mat_nmf<0] <- 0 # negative values should be set to 0 in initial matrix
  239. # output to a list of gene expression profiles (GEP)
  240. per_sample_mat[[i]] <- count_mat_nmf
  241. names(per_sample_mat)[i] <- sample_ls[[i]]
  242. rm(m,m_loged, var_filter, exp_genes, m_proc,count_mat_cent)
  243. #saveRDS(count_mat_nmf, paste("MP/NMF/NMF_mats_GBM/", samples_names[i], ".rds", sep =""))
  244. }
  247. ## Extract sample names for sample list file:
  248. #samples <- list.files("MP/NMF/NMF_mats_GBM/") %>% substri(., pos = 1) %>% str_c(., collapse = "\n")
  249. #write_lines(samples, ("/samples.txt"), sep = "\n")
  254. ###### NMF (Previously run by server. Possible to run per sample below)---------------------------------------------------------------------
  256. #!/usr/bin/env Rscript
  257. #args = commandArgs(trailingOnly = TRUE)
  260. sname <- "MGH258"# enter sample name here
  261. rank_lb <- 2
  262. rank_ub <- 11
  264. m <- readRDS(paste0("MP/NMF/NMF_mats_GBM/", sname, ".rds"))
  265. m <- as.matrix(m)
  266. res <- NMF::nmf(x = m, rank = rank_lb:rank_ub, nrun = 5, method = "snmf/r", .opt = list(debug=F, parallel=F, shared.memory=F, verbose=T))
  269. ###### Generation of NMF + Leiden GBM spatial metaprograms#########
  271. #
  273. #robust_nmf_programs function
  275. robust_nmf_programs <- function(nmf_programs, intra_min = 30, intra_max = 10, inter_filter=T, inter_min = 10) {
  277. # Select NMF programs based on the minimum overlap with other NMF programs from the same sample
  278. intra_intersect <- lapply(nmf_programs, function(z) apply(z, 2, function(x) apply(z, 2, function(y) length(intersect(x,y)))))
  279. intra_intersect_max <- lapply(intra_intersect, function(x) apply(x, 2, function(y) sort(y, decreasing = T)[2]))
  280. nmf_sel <- lapply(names(nmf_programs), function(x) nmf_programs[[x]][,intra_intersect_max[[x]]>=intra_min])
  281. names(nmf_sel) <- names(nmf_programs)
  283. # Select NMF programs based on i) the maximum overlap with other NMF programs from the same sample and
  284. # ii) the minimum overlap with programs from another sample
  285. nmf_sel_unlist <- do.call(cbind, nmf_sel)
  286. inter_intersect <- apply(nmf_sel_unlist , 2, function(x) apply(nmf_sel_unlist , 2, function(y) length(intersect(x,y)))) ## calculating intersection between all programs
  288. final_filter <- NULL
  289. for(i in names(nmf_sel)) {
  290. a <- inter_intersect[grep(i, colnames(inter_intersect), invert = T),grep(i, colnames(inter_intersect))]
  291. b <- sort(apply(a, 2, max), decreasing = T) # for each sample, ranks programs based on their maximum overlap with programs of other samples
  292. if(inter_filter==T) b <- b[b>=inter_min] # selects programs with a maximum intersection of at least 10
  293. if(length(b) > 1) {
  294. c <- names(b[1])
  295. for(y in 2:length(b)) {
  296. if(max(inter_intersect[c,names(b[y])]) <= intra_max) c <- c(c,names(b[y])) # selects programs iteratively from top-down. Only selects programs that have a intersection smaller than 10 with a previously selected programs
  297. }
  298. final_filter <- c(final_filter, c)
  299. } else {
  300. final_filter <- c(final_filter, names(b))
  301. }
  302. }
  303. return(final_filter)
  304. }
  307. # Custom color palette
  309. custom_magma <- c(colorRampPalette(c("white", rev(magma(323, begin = 0.15))[1]))(10), rev(magma(323, begin = 0.18)))
  311. ## Create list of NMF matrics where each sample is an entry
  312. path <- "MP/NMF/out_dir/"
  313. sample_ls <- list.files(path)
  315. ## Create list of NMF matrics where each sample is an entry
  316. prog_genes_ls <- list()
  317. for(i in seq_along(sample_ls)) {
  318. nmf_obj <- readRDS(paste(path, sample_ls[[i]], sep = "/"))
  319. samp_name <- stringr::str_split(sample_ls[[i]], pattern = "_")[[1]][[1]]
  320. nmf_mats <- c()
  321. for(j in names(nmf_obj$fit)) {
  322. get_basis <- basis(nmf_obj$fit[[j]])
  323. colnames(get_basis) <- paste0(samp_name, ".", j, ".", 1:j)
  324. nmf_mats <- cbind(nmf_mats, get_basis)
  325. }
  326. prog_genes_ls[[i]] <- nmf_mats
  327. names(prog_genes_ls)[i] <- paste0(samp_name)
  328. rm(nmf_obj, nmf_mats, get_basis)
  329. }
  330. Genes_nmf_w_basis <- do.call(c, list(prog_genes_ls))
  332. # Find robust NMFs
  333. # get gene programs (top 50 genes by NMF score)
  334. nmf_programs_sig <- lapply(prog_genes_ls, function(x) apply(x, 2, function(y) names(sort(y, decreasing = T))[1:50]))
  336. # for each sample, select robust NMF programs (i.e. observed using different ranks in the same sample), remove redundancy due to multiple ranks, and apply a filter based on the similarity to programs from other samples.
  337. nmf_filter_all <- robust_nmf_programs(nmf_programs_sig, intra_min = 35, intra_max = 10, inter_filter=T, inter_min = 10)
  338. nmf_programs_sig <- lapply(nmf_programs_sig, function(x) x[, is.element(colnames(x), nmf_filter_all),drop=F])
  339. nmf_programs_sig <- do.call(cbind, nmf_programs_sig)
  341. #leiden clusters
  342. all_leiden_programs<-readRDS("MP/all_leiden_programs_v2.rds")
  343. all_leiden_programs <- do.call(cbind, all_leiden_programs)
  344. nmf_programs_sig<-cbind(nmf_programs_sig,all_leiden_programs)
  346. # calculate similarity between programs
  347. nmf_intersect <- apply(nmf_programs_sig , 2, function(x) apply(nmf_programs_sig , 2, function(y) length(intersect(x,y))))
  349. # hierarchical clustering of the similarity matrix
  350. nmf_intersect_hc_all <- hclust(as.dist(50-nmf_intersect), method="average")
  351. nmf_intersect_hc_all <- reorder(as.dendrogram(nmf_intersect_hc_all), colMeans(nmf_intersect))
  352. nmf_intersect <- nmf_intersect[order.dendrogram(nmf_intersect_hc_all), order.dendrogram(nmf_intersect_hc_all)]
  354. nmf_intersect<-readRDS("MP/NMF/nmf_intersect_124.RDS")
  355. nmf_programs_sig<-readRDS("MP/NMF/nmf_programs_sig_124.RDS")
  358. ### use a clustering approach that updates MPs in each iteration
  360. nmf_intersect_KEEP <- nmf_intersect
  361. nmf_programs_sig_KEEP <- nmf_programs_sig
  364. ### Parameters (later change to function form)v1-keep!
  366. Min_intersect_initial <- 12 # the minimal intersection cutoff for defining the Founder NMF program of a cluster
  367. Min_intersect_cluster <- 12 # the minimal intersection cuttof for adding a new NMF to the forming cluster
  368. Min_group_size <- 4 # the minimal group size to consider for defining the Founder_NMF of a MP
  370. Sorted_intersection <- sort(apply(nmf_intersect , 2, function(x) (length(which(x>=Min_intersect_initial))-1) ) , decreasing = TRUE)
  372. Cluster_list <- list() ### Every entry contains the NMFs of a chosec cluster
  373. k <- 1
  374. Curr_cluster <- c()
  375. MP_list <- list()
  377. while (Sorted_intersection[1]>Min_group_size) {
  379. Curr_cluster <- c(Curr_cluster , names(Sorted_intersection[1]))
  381. ### intersection between all remaining NMFs and Genes in MP
  382. Genes_MP <- nmf_programs_sig[,names(Sorted_intersection[1])] # initial genes are those in the first NMF. Genes_MP always has only 50 genes consisting of the current MP
  383. nmf_programs_sig <- nmf_programs_sig[,-match(names(Sorted_intersection[1]) , colnames(nmf_programs_sig))] # remove selected NMF
  384. Intersection_with_Genes_MP <- sort(apply(nmf_programs_sig, 2, function(x) length(intersect(Genes_MP,x))) , decreasing = TRUE) # intersection between all other NMFs and Genes_MP
  385. NMF_history <- Genes_MP # has all genes in all NMFs in the current cluster, for newly defining Genes_MP after adding a new NMF
  387. ### Create gene list - composed of intersecting genes in descending order. Update Curr_cluster each time.
  389. while ( Intersection_with_Genes_MP[1] >= Min_intersect_cluster) { ### Define current cluster
  391. Curr_cluster <- c(Curr_cluster , names(Intersection_with_Genes_MP)[1])
  393. Genes_MP_temp <- sort(table(c(NMF_history , nmf_programs_sig[,names(Intersection_with_Genes_MP)[1]])), decreasing = TRUE) ## Genes_MP is newly defined each time according to all NMFs in the current cluster
  394. Genes_at_border <- Genes_MP_temp[which(Genes_MP_temp == Genes_MP_temp[50])] ### genes with overlap equal to the 50th gene
  396. if (length(Genes_at_border)>1){
  397. ### Sort last genes in Genes_at_border according to maximal NMF gene scores
  398. ### Run over all NMF programs in Curr_cluster and extract NMF scores for each gene
  399. Genes_curr_NMF_score <- c()
  400. for (i in Curr_cluster) {
  401. curr_study <- strsplit(i, "[.]")[[1]][[1]]
  402. Q <- Genes_nmf_w_basis[[curr_study]][ match(names(Genes_at_border),toupper(rownames(Genes_nmf_w_basis[[curr_study]])))[!is.na(match(names(Genes_at_border),toupper(rownames(Genes_nmf_w_basis[[curr_study]]))))] ,i]
  403. #names(Q) <- names(Genes_at_border[!is.na(match(names(Genes_at_border),toupper(rownames(Genes_nmf_w_basis[[curr_study]]))))]) ### sometimes when adding genes the names do not appear
  404. Genes_curr_NMF_score <- c(Genes_curr_NMF_score, Q )
  405. }
  406. Genes_curr_NMF_score_sort <- sort(Genes_curr_NMF_score , decreasing = TRUE)
  407. Genes_curr_NMF_score_sort <- Genes_curr_NMF_score_sort[unique(names(Genes_curr_NMF_score_sort))] ## take only the maximal score of each gene - which is the first entry after sorting
  409. Genes_MP_temp <- c(names(Genes_MP_temp[which(Genes_MP_temp > Genes_MP_temp[50])]) , names(Genes_curr_NMF_score_sort))
  411. } else {
  412. Genes_MP_temp <- names(Genes_MP_temp)[1:50]
  413. }
  415. NMF_history <- c(NMF_history , nmf_programs_sig[,names(Intersection_with_Genes_MP)[1]])
  416. Genes_MP <- Genes_MP_temp[1:50]
  418. nmf_programs_sig <- nmf_programs_sig[,-match(names(Intersection_with_Genes_MP)[1] , colnames(nmf_programs_sig))] # remove selected NMF
  420. Intersection_with_Genes_MP <- sort(apply(nmf_programs_sig, 2, function(x) length(intersect(Genes_MP,x))) , decreasing = TRUE) # intersection between all other NMFs and Genes_MP
  422. }
  424. Cluster_list[[paste0("Cluster_",k)]] <- Curr_cluster
  425. MP_list[[paste0("MP_",k)]] <- Genes_MP
  426. k <- k+1
  429. nmf_intersect <- nmf_intersect[-match(Curr_cluster,rownames(nmf_intersect) ) , -match(Curr_cluster,colnames(nmf_intersect) ) ] # remove current chosen cluster
  431. Sorted_intersection <- sort(apply(nmf_intersect , 2, function(x) (length(which(x>=Min_intersect_initial))-1) ) , decreasing = TRUE) ### Sort intersection of remaining NMFs not included in any of the previous clusters
  433. Curr_cluster <- c()
  434. print(dim(nmf_intersect)[2])
  435. }
  437. #### ***** Sort Jaccard similarity plot according to new clusters:
  439. inds_sorted <- c()
  441. for (j in 1:length(Cluster_list)){
  443. inds_sorted <- c(inds_sorted , match(Cluster_list[[j]] , colnames(nmf_intersect_KEEP)))
  445. }
  446. inds_new <- c(inds_sorted , which(is.na( match(1:dim(nmf_intersect_KEEP)[2],inds_sorted)))) ### combine inds in clusters with inds without clusters
  449. # plot re-ordered similarity matrix heatmap
  450. nmf_intersect_meltI_NEW <- reshape2::melt(nmf_intersect_KEEP[inds_new,inds_new])
  452. custom_magma <- c(colorRampPalette(c("white", rev(magma(323, begin = 0.15))[1]))(10), rev(magma(323, begin = 0.18)))
  455. ggplot(data = nmf_intersect_meltI_NEW, aes(x=Var1, y=Var2, fill=100*value/(100-value), color=100*value/(100-value))) +
  456. geom_tile() +
  457. scale_color_gradient2(limits=c(2,25), low=custom_magma[1:111], mid =custom_magma[112:222], high = custom_magma[223:333], midpoint = 13.5, oob=squish, name="Similarity\n(Jaccard index)") +
  458. scale_fill_gradient2(limits=c(2,25), low=custom_magma[1:111], mid =custom_magma[112:222], high = custom_magma[223:333], midpoint = 13.5, oob=squish, name="Similarity\n(Jaccard index)") +
  459. theme( axis.ticks = element_blank(), panel.border = element_rect(fill=F), panel.background = element_blank(), axis.line = element_blank(), axis.text = element_text(size = 11), axis.title = element_text(size = 12), legend.title = element_text(size=11), legend.text = element_text(size = 10), legend.text.align = 0.5, legend.justification = "bottom") +
  460. theme(axis.title.x=element_blank(), axis.text.x=element_blank(), axis.ticks.x=element_blank()) +
  461. theme(axis.title.y=element_blank(), axis.text.y=element_blank(), axis.ticks.y=element_blank()) +
  462. guides(fill = guide_colourbar(barheight = 4, barwidth = 1))
  464. ggplot(data = nmf_intersect_meltI_NEW, aes(x=Var1, y=Var2, fill=100*value/(100-value), color=100*value/(100-value))) +
  465. geom_tile() +
  466. scale_color_gradient2(limits=c(2,30), low=custom_magma[1:111], mid =custom_magma[112:222], high = custom_magma[223:333], midpoint = 16, oob=squish, name="Similarity\n(Jaccard index)") +
  467. scale_fill_gradient2(limits=c(2,30), low=custom_magma[1:111], mid =custom_magma[112:222], high = custom_magma[223:333], midpoint = 16, oob=squish, name="Similarity\n(Jaccard index)") +
  468. theme( axis.ticks = element_blank(), panel.border = element_rect(fill=F), panel.background = element_blank(), axis.line = element_blank(), axis.text = element_text(size = 11), axis.title = element_text(size = 12), legend.title = element_text(size=11), legend.text = element_text(size = 10), legend.text.align = 0.5, legend.justification = "bottom") +
  469. theme(axis.title.x=element_blank(), axis.text.x=element_blank(), axis.ticks.x=element_blank()) +
  470. theme(axis.title.y=element_blank(), axis.text.y=element_blank(), axis.ticks.y=element_blank()) +
  471. guides(fill = guide_colourbar(barheight = 4, barwidth = 1))
  474. MP <- do.call(cbind, MP_list)
  476. names(MP_list)=c("Neuron","Vasc","MES.Hyp","Mac","OPC.AC","Oligo","LQ.Chromatin.reg","MES","Prolif.Metab","MES.Ast","Reactive.Ast","NPC","Inflammatory.Mac")
  479. ###### RECLUSTERING/HEIRARCHICAL CLUSTERING, EXTENDED MPs, AND SUBCLUSTER PROGRAMS ################
  480. programs_ls<-readRDS("MP/NMF/programs_ls_124.rds") #this is "nmf_programs_sig"
  481. Cluster_list<-readRDS("MP/NMF/Cluster_list_124.rds")
  482. str(Cluster_list)
  483. MP_list<-readRDS("MP/NMF/combined_gbm_metaprograms_raw_124.rds")
  486. #reclustering OPC.AC into 2 clusters
  487. opc.ac <- Cluster_list[[5]]
  489. mat2list <- function(x) {
  490. stopifnot(is.matrix(x))
  491. lapply(seq_len(ncol(x)), function(i) x[, i])
  492. }
  493. program_ls<-mat2list(programs_ls)
  494. names(program_ls)<-colnames(programs_ls)
  495. opc.ac_mat <- program_ls[names(program_ls) %in% opc.ac]
  497. jac_mat <- scalop::jaccard(opc.ac_mat)
  498. hc <- hclust(dist(1 - jac_mat), method = "average")
  499. jac_plot <- melt(jac_mat[hc$order, hc$order])
  500. #summary(jac_plot$value)
  502. p1 <- ggplot(data = jac_plot, aes(x = Var1, y = Var2, fill = value)) + geom_raster() +
  503. scale_fill_gradient2(limits = c(0, 0.4), low = rev(magma(323, begin = 0.15)), mid = "antiquewhite", high = rev(magma(323, begin = 0.18)), midpoint = 0 , oob = squish, name = "Pearson\nCorrelation") +
  504. theme(axis.ticks = element_blank(), panel.background = element_rect(fill = "white"), axis.line = element_blank(),axis.title = element_text(size=24), legend.text.align = 0.5) +
  505. scale_x_discrete(name = "\nPrograms", breaks = colnames(jac_mat)[hc$order][seq(5, ncol(jac_mat), by = 5)], labels = seq(5, ncol(jac_mat), by = 5)) +
  506. scale_y_discrete(name = "\nPrograms", breaks = colnames(jac_mat)[hc$order][seq(5, ncol(jac_mat), by = 5)], labels = seq(5, ncol(jac_mat), by = 5))
  508. prog_ordered <- colnames(jac_mat[hc$order, hc$order])
  509. plot(hc)
  510. rect.hclust(hc , k = 2, border = 2:6)
  511. clusterCut <- stats::cutree(tree = hc, k = 2)
  513. metaprog_df <- data.frame(row.names = prog_ordered)
  514. metaprog_df$cluster <- (clusterCut[prog_ordered])
  515. metaprog_df$cells <- prog_ordered
  516. metaprog_df$cluster <- as.factor(metaprog_df$cluster)
  518. metaprog_annotation_plot <- ggplot(metaprog_df, aes(x = factor(cells, levels = cells), y = 1)) +
  519. geom_raster(aes(fill = cluster)) + theme(legend.position = "bottom", axis.ticks = element_blank(),axis.title = element_text(size=24), panel.background = element_rect(fill = "white"), axis.text = element_blank(), axis.line = element_blank(), axis.text.x = element_blank()) +
  520. labs(x = "", y = "", fill = "")
  522. egg::ggarrange(p1, metaprog_annotation_plot, ncol = 1, nrow = 2, heights = c(40, 2))
  524. clust_list_ac_opc <- list(names(clusterCut[clusterCut == 1]), names(clusterCut[clusterCut == 2])) # use this ordering to reorder cluster 5 in main MP heatmap
  525. mp_freqs <- sapply(clust_list_ac_opc, function(k) sort(table(unlist(opc.ac_mat[k])), decreasing = T), simplify = F)
  526. opc.ac_metaprograms <- sapply(mp_freqs, function(tab) head(names(tab)[tab >= 3], 50), simplify = F)
  527. names(opc.ac_metaprograms)<-c("AC","OPC")
  530. #dividing low quality from chromatin.reg cluster
  531. chromatin <- Cluster_list[[7]]
  532. mat2list <- function(x) {
  533. stopifnot(is.matrix(x))
  534. lapply(seq_len(ncol(x)), function(i) x[, i])
  535. }
  536. program_ls<-mat2list(programs_ls)
  537. names(program_ls)<-colnames(programs_ls)
  538. chromatin_mat <- program_ls[names(program_ls) %in% chromatin]
  540. jac_mat <- scalop::jaccard(chromatin_mat)
  541. hc <- hclust(dist(1 - jac_mat), method = "average")
  542. jac_plot <- melt(jac_mat[hc$order, hc$order])
  543. #summary(jac_plot$value)
  545. p1 <- ggplot(data = jac_plot, aes(x = Var1, y = Var2, fill = value)) + geom_raster() +
  546. scale_fill_gradient2(limits = c(0, 0.4), low = rev(magma(323, begin = 0.15)), mid = "antiquewhite", high = rev(magma(323, begin = 0.18)), midpoint = 0 , oob = squish, name = "Pearson\nCorrelation") +
  547. theme(axis.ticks = element_blank(), panel.background = element_rect(fill = "white"), axis.line = element_blank(),axis.title = element_text(size=24), legend.text.align = 0.5) +
  548. scale_x_discrete(name = "\nPrograms", breaks = colnames(jac_mat)[hc$order][seq(5, ncol(jac_mat), by = 5)], labels = seq(5, ncol(jac_mat), by = 5)) +
  549. scale_y_discrete(name = "\nPrograms", breaks = colnames(jac_mat)[hc$order][seq(5, ncol(jac_mat), by = 5)], labels = seq(5, ncol(jac_mat), by = 5))
  552. prog_ordered <- colnames(jac_mat[hc$order, hc$order])
  553. plot(hc)
  554. rect.hclust(hc , k = 3, border = 2:6)
  555. clusterCut <- stats::cutree(tree = hc, k = 3)
  557. metaprog_df <- data.frame(row.names = prog_ordered)
  558. metaprog_df$cluster <- (clusterCut[prog_ordered])
  559. metaprog_df$cells <- prog_ordered
  560. metaprog_df$cluster <- as.factor(metaprog_df$cluster)
  562. metaprog_annotation_plot <- ggplot(metaprog_df, aes(x = factor(cells, levels = cells), y = 1)) +
  563. geom_raster(aes(fill = cluster)) + theme(legend.position = "bottom", axis.ticks = element_blank(),panel.background = element_rect(fill = "white"), axis.text = element_blank(), axis.line = element_blank(), axis.text.x = element_blank(),axis.title = element_text(size=16)) +
  564. labs(x = "", y = "", fill = "")
  566. egg::ggarrange(p1, metaprog_annotation_plot, ncol = 1, nrow = 2, heights = c(40, 2))
  568. clust_list <- list(names(clusterCut[clusterCut == 1]), names(clusterCut[clusterCut == 2]),names(clusterCut[clusterCut == 3]))
  569. mp_freqs <- sapply(clust_list, function(k) sort(table(unlist(chromatin_mat[k])), decreasing = T), simplify = F)
  570. chromatin_metaprograms <- sapply(mp_freqs, function(tab) head(names(tab)[tab >= 3], 50), simplify = F)
  571. names(chromatin_metaprograms)<-c("LQ1","chromatin_clean","LQ2")
  572. chromatin_metaprogram_clean<-chromatin_metaprograms$chromatin_clean
  575. clust_list2 <- list(names(clusterCut[clusterCut == 1]), names(clusterCut[clusterCut == 3]),names(clusterCut[clusterCut == 2]))
  576. #this is the order to use for the final MP heatmap
  577. clust_list_chromatin <- c(names(clusterCut[clusterCut == 1]), names(clusterCut[clusterCut == 2]),names(clusterCut[clusterCut == 3]))
  579. ###### MPs after OP/AC split and LQ/chromatin.reg split#######
  580. MP_list2<-MP_list[-c(5,7)]
  582. MP_list2[["chromatin.reg"]] <- chromatin_metaprogram_clean
  583. MP_list2[["OPC"]] <- opc.ac_metaprograms$OPC
  584. MP_list2[["AC"]] <- opc.ac_metaprograms$AC
  587. ###### SPOT ASSIGNMENTS TO CLEANED METAPROGRAMS##########
  589. # generate normalized exp matrices
  590. file_paths <- list.files(path ="general/exp_mats_GBM", pattern = "\\.rds", full.names = TRUE)
  591. sample_ls <- gsub(pattern = "\\.rds$", replacement = "", x = basename(file_paths))
  592. sample_ls->samples_names
  593. sample_ls->samples
  594. per_sample_mat <- lapply(file_paths, readRDS)
  596. for (i in seq_along(per_sample_mat)){
  597. m <- as.matrix(per_sample_mat[[i]])
  598. m <- m[-grep("^MT-|^RPL|^RPS", rownames(m)), ]
  599. if(min(colSums(m)) == 0){m <- m[, colSums(m) != 0]}
  600. scaling_factor <- 1000000/colSums(m)
  601. m_CPM <- sweep(m, MARGIN = 2, STATS = scaling_factor, FUN = "*")
  602. m_loged <- log2(1 + (m_CPM/10))
  604. # removing genes with zero variance across all cells
  605. var_filter <- apply(m_loged, 1, var)
  606. m_proc <- m_loged[var_filter != 0, ]
  607. # filtering out lowly expressed genes
  608. exp_genes <- rownames(m_proc)[(rowMeans(m_proc) > 0.4)]
  609. m_proc <- m_proc[exp_genes, ]
  611. # output to a list of gene expression profiles (GEP)
  612. per_sample_mat[[i]] <- m_proc
  613. names(per_sample_mat)[i] <- sample_ls[[i]]
  614. rm(m,m_loged, var_filter, exp_genes, m_proc)
  615. }
  617. #generate a list with the filtered exp matrix for each sample, i.e. m_proc<-(per_sample_mat[[1]])
  618. score_mat <- lapply(c(1:length(per_sample_mat)), function(i){
  619. m_proc<-per_sample_mat[[i]]
  620. metaprograms_gene_list <- readRDS("MP/clean_spatial_gbm_metaprograms_124.rds")
  621. signatures <- scalop::sigScores(m_proc, metaprograms_gene_list, expr.center = TRUE, conserved.genes = 0.5)
  623. spot_scores <- data.frame(spot_names = rownames(signatures))
  624. spot_scores$Neuron <- signatures$Neuron
  625. spot_scores$Vasc<- signatures$Vasc
  626. spot_scores$MES.Hyp <- signatures$MES.Hyp
  627. spot_scores$Mac <- signatures$Mac
  628. spot_scores$Oligo <- signatures$Oligo
  629. spot_scores$MES <- signatures$MES
  630. spot_scores$Prolif.Metab <- signatures$Prolif.Metab
  631. spot_scores$MES.Ast <- signatures$MES.Ast
  632. spot_scores$Reactive.Ast <- signatures$Reactive.Ast
  633. spot_scores$NPC <- signatures$NPC
  634. spot_scores$Inflammatory.Mac <- signatures$Inflammatory.Mac
  635. spot_scores$chromatin.reg <- signatures$chromatin.reg
  636. spot_scores$OPC <- signatures$OPC
  637. spot_scores$AC <- signatures$AC
  638. return(spot_scores)
  639. })
  641. names(score_mat) <- sample_ls
  643. for (i in seq_along(score_mat)){
  644. score_df <- as.data.frame(score_mat[[i]])
  645. score_df <-column_to_rownames(score_df, 'spot_names')
  646. maxcol_meta<-maxcol_strict(score_df)
  647. maxcol_meta<-stack(maxcol_meta)
  648. setnames(maxcol_meta,2,"spot_type_meta_new")
  649. setnames(maxcol_meta,1,"SpotID")
  650. }

Module1_MP.R at commit f38f9b6, no license · at the source

Overview

Authors: Genaro R Villa1, Paolo Alimonti1, Joseph S Toker2, Raziye Piranlioglu1, Mikayla A Karkoski1, Debora Mazzetti1, Reda Ben Mrid3, Sara El Guendouzi3, Alexa Lauinger1,4, Andrew N Chiocca1, Rachid El Fatimy3, E Antonio Chiocca1, Marco Mineo1
ORCID iDs: Genaro R Villa, Paolo Alimonti, Raziye Piranlioglu, Reda Ben Mrid, Alexa Lauinger, E Antonio Chiocca, Marco Mineo
  1. Harvey W. Cushing Neuro-oncology Laboratories, Department of Neurosurgery, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA USA
  2. Harvard Medical School, Boston, MA USA
  3. Faculty of Medical Sciences, UM6P Hospitals, Mohammed VI Polytechnic University, Benguerir, Morocco
  4. Carle Illinois College of Medicine, Urbana, IL USA
Institutions: Brigham and Women's Hospital (United States); Harvard University (United States); Université Mohammed VI Polytechnique (Morocco)
Journal: Cell death & disease, volume 17, issue 1, article 352
Dates: received 17 June 2025; accepted 6 March 2026; published online 24 March 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1038/s41419-026-08576-6 · PMID 41872153 · PMCID PMC13039110 · OpenAlex W7140234994
Open access: gold, a free copy (OpenAlex)
Status: code verified (how the code is found and verified)
Categories: human (organism), mouse (organism), other condition (population), cellular / molecular (subfield)
Methods: Statistics, Smoothing, state filtering, decompositions, Evoked potentials, Connectivity
Keywords: CNS cancer, RNA splicing
MeSH: Brain Neoplasms*, Genes, cdc*, Glioblastoma*, Heterogeneous-Nuclear Ribonucleoprotein Group F-H*, RNA Splicing*, Animals, Cell Cycle, Cell Line, Tumor, Cell Proliferation, Disease Progression, G2 Phase Cell Cycle Checkpoints, Gene Expression Regulation, Neoplastic, Humans, Mice (* major topic)
Topic: RNA Research and Splicing (Molecular Biology, Biochemistry, Genetics and Molecular Biology), according to OpenAlex
Funding: Neurosurgery Research and Education Foundation; U.S. Department of Health &amp; Human Services | NIH | National Institute of General Medical Sciences (T32GM144273); NIGMS NIH HHS (T32 GM144273); U.S. Department of Health & Human Services | NIH | National Institute of General Medical Sciences (NIGMS) (T32GM144273)
Citations: cited by 1 paper (Europe PMC); 41 references in the paper

Abstract

Although glioblastoma (GBM) harbors multiple genetic abnormalities leading to cell cycle deregulation, a functional mitotic checkpoint is essential to prevent mitotic catastrophe and tumor cell death. Here, we identify the RNA-binding protein HNRNPH1 as a key post-transcriptional modulator of G2/M checkpoint-associated genes in GBM. HNRNPH1 is overexpressed in malignant cells, especially in the neural- and oligodendrocyte-progenitor-like state, and its expression levels are higher in non-hypoxic regions of the tumor. Knocking out HNRNPH1 causes aberrant splicing and downregulation of several genes involved in cell division. These molecular alterations are associated with G2/M cell cycle arrest, reduced cell proliferation, abnormal cell morphology, and increased nuclear fragmentation. Silencing HNRNPH1 in vivo inhibits the tumor growth of patient-derived GBM cell-originated intracranial xenografts and has significant survival benefits. Together, our results show the critical importance of HNRNPH1 in cell cycle progression and tumor growth, potentially impacting the development of novel strategies to treat GBM.

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

Repository

Its files are read in the Code ↔ Paper reader above, with 2 matches between paragraphs and lines of code.

tiroshlab/Spatial_Glioma

License: none: the authors keep all their rights
State: the link answers, verified on 30 September 2026
Evidence: files inventoried
Commit: f38f9b60ddbeefd9739e1483cbc66b2650f931d3, 6 January 2025
Languages: R (6)
Size: 7 files, 6 scripts
Software Heritage: not archived
Found in: the text, “Analysis of spatial transcriptomics data”
Holds: README
Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: tidyverse (4 files), patchwork (3 files), data.table (2 files), ggpubr (2 files), igraph (2 files), reshape2 (2 files), circlize (1 file), ComplexHeatmap (1 file), ggplot2 (1 file), reticulate (1 file), Seurat (1 file), SingleCellExperiment (1 file)
Availability: 1 check, the latest on 30 September 2026: the link answers
  • 30 September 2026: the link answers
7 files
At the source: github.com/tiroshlab/Spatial_Glioma

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;
  • 6 scripts, each with its path and the digest of its content;
  • 2 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 availability

The datasets generated and/or analyzed during the current study are available in the Gene Expression Omnibus (GEO), GSE294242, GSE294345, and GSE137489 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE137489).

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, 30 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 13 authors, 2 keywords, 14 MeSH terms, 4 funders, 41 references.

Cite

This paper

Villa, G. R., Alimonti, P., Toker, J. S., Piranlioglu, R., Karkoski, M. A., Mazzetti, D., Ben Mrid, R., El Guendouzi, S., Lauinger, A., Chiocca, A. N., El Fatimy, R., Chiocca, E. A., & Mineo, M. (2026). HNRNPH1 drives glioblastoma progression by regulating the splicing of cell cycle genes. Cell death & disease, 17(1), 352. https://doi.org/10.1038/s41419-026-08576-6

BibTeX

@article{villa2026hnrnph1,
author = {Villa, Genaro R and Alimonti, Paolo and Toker, Joseph S and Piranlioglu, Raziye and Karkoski, Mikayla A and Mazzetti, Debora and Ben Mrid, Reda and El Guendouzi, Sara and Lauinger, Alexa and Chiocca, Andrew N and El Fatimy, Rachid and Chiocca, E Antonio and Mineo, Marco},
title = {{HNRNPH1 drives glioblastoma progression by regulating the splicing of cell cycle genes}},
journal = {Cell death \& disease},
year = {2026},
month = mar,
volume = {17},
number = {1},
pages = {352},
publisher = {Nature Publishing Group},
issn = {2041-4889},
doi = {10.1038/s41419-026-08576-6},
url = {https://doi.org/10.1038/s41419-026-08576-6},
pmid = {41872153},
pmcid = {PMC13039110}
}

RIS

TY - JOUR
AU - Villa, Genaro R
AU - Alimonti, Paolo
AU - Toker, Joseph S
AU - Piranlioglu, Raziye
AU - Karkoski, Mikayla A
AU - Mazzetti, Debora
AU - Ben Mrid, Reda
AU - El Guendouzi, Sara
AU - Lauinger, Alexa
AU - Chiocca, Andrew N
AU - El Fatimy, Rachid
AU - Chiocca, E Antonio
AU - Mineo, Marco
TI - HNRNPH1 drives glioblastoma progression by regulating the splicing of cell cycle genes
T2 - Cell death & disease
J2 - Cell Death Dis
PY - 2026
DA - 2026/03/24
VL - 17
IS - 1
SP - 352
SN - 2041-4889
PB - Nature Publishing Group
DO - 10.1038/s41419-026-08576-6
UR - https://doi.org/10.1038/s41419-026-08576-6
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s41419-026-08576-6",
"type": "article-journal",
"title": "HNRNPH1 drives glioblastoma progression by regulating the splicing of cell cycle genes",
"container-title": "Cell death & disease",
"author": [
{
"family": "Villa",
"given": "Genaro R"
},
{
"family": "Alimonti",
"given": "Paolo"
},
{
"family": "Toker",
"given": "Joseph S"
},
{
"family": "Piranlioglu",
"given": "Raziye"
},
{
"family": "Karkoski",
"given": "Mikayla A"
},
{
"family": "Mazzetti",
"given": "Debora"
},
{
"family": "Ben Mrid",
"given": "Reda"
},
{
"family": "El Guendouzi",
"given": "Sara"
},
{
"family": "Lauinger",
"given": "Alexa"
},
{
"family": "Chiocca",
"given": "Andrew N"
},
{
"family": "El Fatimy",
"given": "Rachid"
},
{
"family": "Chiocca",
"given": "E Antonio"
},
{
"family": "Mineo",
"given": "Marco"
}
],
"container-title-short": "Cell Death Dis",
"volume": "17",
"issue": "1",
"page": "352",
"DOI": "10.1038/s41419-026-08576-6",
"PMID": "41872153",
"PMCID": "PMC13039110",
"ISSN": "2041-4889",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s41419-026-08576-6",
"language": "en",
"issued": {
"date-parts": [
[
2026,
3,
24
]
]
}
}

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

Similar papers

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

[1] doi:10.1016/j.celrep.2024.115149 [code]
Single-nucleus and spatial landscape of the sub-ventricular zone in human glioblastoma.
Journal: Cell reports
In common: reticulate, SingleCellExperiment, circlize, 9 other tools, other condition, cellular / molecular, 3 references
[2] doi:10.1038/s41586-026-10214-2 [code]
Multidimensional profiling of heterogeneity in supratentorial ependymomas.
Journal: Nature
In common: reticulate, SingleCellExperiment, circlize, 9 other tools, other condition, mouse, 2 references
[3] doi:10.1038/s41586-023-06623-2 [code]
Clinical trial links oncolytic immunoactivation to survival in glioblastoma.
Journal: Nature
In common: circlize, ComplexHeatmap, ggpubr, 2 other tools, other condition, 2 authors
[4] doi:10.1016/j.xcrm.2026.102766 [code]
A longitudinal single-cell and spatial multiomic atlas of pediatric high-grade glioma.
Journal: Cell reports. Medicine
In common: SingleCellExperiment, circlize, ComplexHeatmap, 8 other tools, other condition, cellular / molecular, 2 references
[5] doi:10.1038/s41586-026-10612-6 [code]
Acquired genetic and cell-state changes in IDH-mutant glioma progression.
Journal: Nature
In common: SingleCellExperiment, ComplexHeatmap, igraph, 7 other tools, other condition, cellular / molecular, 3 references
[6] doi:10.1038/s41591-024-02969-w [code]
A prognostic neural epigenetic signature in high-grade glioma.
Journal: Nature medicine
In common: reticulate, SingleCellExperiment, igraph, 7 other tools, other condition, mouse, 3 references
[7] doi:10.1016/j.isci.2022.105276 [code]
Integrative genomic analysis facilitates precision strategies for glioblastoma treatment.
Journal: iScience
In common: reticulate, SingleCellExperiment, igraph, 7 other tools, other condition, 3 references
[8] doi:10.1172/jci.insight.179853 [code]
Spatial analysis of recurrent glioblastoma reveals perivascular niche organization.
Journal: JCI insight
In common: reticulate, SingleCellExperiment, ComplexHeatmap, 7 other tools, other condition, cellular / molecular, 2 references
[9] doi:10.1186/s13059-024-03407-3 [code]
Spatiotemporal modeling reveals high-resolution invasion states in glioblastoma.
Journal: Genome biology
In common: circlize, ComplexHeatmap, igraph, 6 other tools, other condition, mouse, 3 references
[10] doi:10.3390/biomedicines13061304 [code]
Multi-Transcriptomic Analysis Reveals GSC-Driven MES-Like Differentiation via EMT in GBM Cell-Cell Communication.
Journal: Biomedicines
In common: SingleCellExperiment, igraph, Seurat, 4 other tools, NCBI GEO GSE131928, other condition, cellular / molecular, 2 references

Contribute

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

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

Add the badge to the code's README

OSCR: the badge of this paper

If you are an author or a maintainer of this paper's code, you can add this badge. It says what OSCR knows of the paper (its code, its validated tracing map) and links the code to this page. Paste one of these snippets under the README's title. Keep the two marks around it, so that the badge can be replaced in place and is never added twice. OSCR changes nothing in your repository.

Markdown
<!-- oscr-badge:start -->
[![OSCR](https://openscicode.org/badge/doi/10.1038/s41419-026-08576-6.svg)](https://openscicode.org/badge/doi/10.1038/s41419-026-08576-6)
<!-- oscr-badge:end -->
reStructuredText
.. oscr-badge:start

.. image:: https://openscicode.org/badge/doi/10.1038/s41419-026-08576-6.svg
   :target: https://openscicode.org/badge/doi/10.1038/s41419-026-08576-6
   :alt: OSCR

.. oscr-badge:end
HTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.1038/s41419-026-08576-6"><img src="https://openscicode.org/badge/doi/10.1038/s41419-026-08576-6.svg" alt="OSCR"></a>
<!-- oscr-badge:end -->

On GitHub, open the README, choose Edit (the pencil), paste the snippet under the title, then "Commit changes".

If you choose to, you can have the badge kept up to date for you, with oscr badge --write from the command line or with the GitHub Action in your repository's workflows. Every other form and option (a card, AsciiDoc, LaTeX, shields.io) is on the badge 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: code errors and mismatches between the code and the paper about this paper and its code, from readers and its authors, beside the paper and its tracing map.

Report a mismatch between the code and the paper · Report a code error (signed in)

Reproductions: reports from readers who ran the authors' code and did not get the paper's result, what they ran, with which environment, commit and data.

Report a reproduction failure (signed in)

Versions of its code: the releases of its code tied to a version of this paper (the preprint, the accepted manuscript, the version of record…), each with the tracing map versioned with it.

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