OSCR

ScQCenrich enables multi-metric quality control for single-cell RNA sequencing.

Code ↔ Paper

13 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 13 matches
  1. [1] § Methods › scQCenrich workflow ↔ R/wrapper.R, lines 3–68 · score 0.94 · enrichment_plots, report_html, run_qc_pipeline, rescue_mode, low quality cells, qc_strength
  2. [2] § Methods › QC burden scoring and removal rules ↔ R/outliers.R, lines 285–333 · score 0.93 · avoid double, remove_quantile, tailed deviation, MALAT1 term, nFeature, pctMT
  3. [3] § Methods › scQCenrich workflow ↔ inst/benchmarking.R, lines 456–538 · score 0.92 · enrichment_plots, qc_report.html, run_qc_pipeline, rescue_mode, qc_strength, qc_outputs
  4. [4] § Methods › Preliminary annotation and clustering ↔ inst/benchmarking.R, lines 86–127 · score 0.91 · FindClusters, FindNeighbors, RunPCA, ScaleData, FindVariableFeatures, NormalizeData
  5. [5] § Methods › Model-based low-quality detection ↔ R/outliers.R, lines 212–283 · score 0.90 · minPts, stress_score, Mclust fails, nFeature, pctMT, QC scoring
  6. [6] § Methods › Preliminary annotation and clustering ↔ R/enrichment.R, lines 82–126 · score 0.90 · FindClusters, FindNeighbors, RunPCA, ScaleData, FindVariableFeatures, NormalizeData
  7. [7] § Methods › Coherence-based rescue ↔ R/rescue.R, lines 103–146 · score 0.90 · available metric gates, Intronic gating, median intronic, rescue score, pass fraction, healthy
  8. [8] § Methods › QC burden scoring and removal rules ↔ R/filter.R, lines 46–99 · score 0.89 · remove_quantile, qc_strength, nFeature, pctMT, stress score, intronic fraction
  9. [9] § Methods › QC metrics calculation ↔ scripts/build_figure6_report_sources.R, lines 593–654 · score 0.81 · unspliced fraction, MALAT1 fraction, intronic fraction, stress score, detected genes, heat
  10. [10] § Methods › QC metrics calculation ↔ R/plots.R, lines 242–312 · score 0.80 · unspliced fraction, MALAT1 fraction, intronic fraction, stress score, detected genes, ratio
  11. [11] § Methods › Default safeguards and final QC states ↔ R/filter.R, lines 397–481 · score 0.68 · pre rescue, Borderline cells, removed cell, demoted, cap, extreme
  12. [12] § Methods › Automated reporting ↔ scripts/build_figure6_report_sources.R, lines 593–654 · score 0.67 · MALAT1 fraction, intronic fraction, stress score, detected genes, GO, mt
  13. [13] § Methods › Splice-aware input generation and comparator methods ↔ inst/benchmarking.R, lines 349–410 · score 0.63 · DropletQC, nuclear fraction, baselines, tuned, intact, unspliced

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 · 538 lines · 19 KB · other · 3 matches

  4. # install if missing ------------------------------------------------------
  6. # 1. Increase timeout for large Bioconductor/GitHub downloads
  7. options(timeout = 600)
  9. # 2. Define the packages by source
  10. cran_pkgs <- c("Seurat", "Matrix", "dplyr", "tidyr", "ggplot2", "cluster", "lme4", "remotes")
  11. bioc_pkgs <- c("scater", "SingleCellExperiment", "scuttle", "miQC", "scDblFinder",
  12. "celda", "SingleR", "celldex")
  13. github_pkgs <- c("DropletQC" = "powellgenomicslab/DropletQC")
  15. # 3. Ensure BiocManager and remotes are installed first
  16. if (!requireNamespace("BiocManager", quietly = TRUE)) install.packages("BiocManager")
  17. if (!requireNamespace("remotes", quietly = TRUE)) install.packages("remotes")
  19. # 4. Install missing CRAN packages
  20. missing_cran <- cran_pkgs[!(cran_pkgs %in% installed.packages()[, "Package"])]
  21. if (length(missing_cran)) install.packages(missing_cran)
  23. # 5. Install missing Bioconductor packages
  24. missing_bioc <- bioc_pkgs[!(bioc_pkgs %in% installed.packages()[, "Package"])]
  25. if (length(missing_bioc)) BiocManager::install(missing_bioc, ask = FALSE)
  27. # 6. Install missing GitHub packages (specifically DropletQC)
  28. if (!requireNamespace("DropletQC", quietly = TRUE)) {
  29. remotes::install_github(github_pkgs["DropletQC"])
  30. }
  32. # 7. Load all libraries
  33. all_pkgs <- c(cran_pkgs, bioc_pkgs, names(github_pkgs))
  34. lapply(all_pkgs, library, character.only = TRUE)
  37. # load lib ----------------------------------------------------------------
  41. library(scater) # addPerCellQCMetrics
  42. library(Seurat)
  43. library(SingleCellExperiment)
  44. library(scuttle)
  45. library(miQC) # mixtureModel, filterCells
  46. library(scDblFinder) # doublets
  47. library(celda) # DecontX
  48. library(SingleR)
  49. library(celldex) # biology annotation
  50. library(Matrix)
  51. library(dplyr)
  52. library(tidyr)
  53. library(ggplot2)
  54. library(cluster) # silhouette
  55. library(lme4) # mixed model
  56. library(DropletQC) # dropletqc
  58. options(timeout = 600)
  61. # Data path-----
  63. raw_counts_dir <- "path/to/filtered_feature_bc_matrix/"
  64. spliced_rds <- "path/to/testloom.rds"
  65. # human or mouse
  66. organism <- "human"
  67. organ <- c("Lung", "Blood")
  70. # preprocessing -------------------------------------------------------
  71. seurat_std <- CreateSeuratObject(counts = Read10X(raw_counts_dir), project = "QC_compare")
  73. seurat_sp <- readRDS(spliced_rds)
  75. # Helper: normalize cell barcodes to a comparable format
  76. .normalize_cells <- function(cells) {
  77. # drop sample prefix if present
  78. cells <- sub(".*:", "", cells)
  79. # drop trailing x if present
  80. cells <- sub("x$", "", cells)
  81. # drop trailing -1 if present
  82. cells <- sub("-1$", "", cells)
  83. cells
  84. }
  86. # Suppose you have:
  87. # seurat_std = your "standard" Seurat object (RNA assay)
  88. # seurat_sp = your velocity-style Seurat object (spliced/unspliced)
  90. # Normalize cell IDs
  91. std_ids <- .normalize_cells(colnames(seurat_std))
  92. sp_ids <- .normalize_cells(colnames(seurat_sp))
  94. # Find intersection
  95. common <- intersect(std_ids, sp_ids)
  96. message("Found ", length(common), " overlapping cells.")
  98. # Sample cells if needed (or fewer if less available)
  99. set.seed(123) # for reproducibility
  100. keep_ids <- sample(common, min(50000, length(common)))
  102. # Map back to original names in each object
  103. std_keep <- colnames(seurat_std)[.normalize_cells(colnames(seurat_std)) %in% keep_ids]
  104. sp_keep <- colnames(seurat_sp)[.normalize_cells(colnames(seurat_sp)) %in% keep_ids]
  106. # Subset
  107. seurat_std_small <- subset(seurat_std, cells = std_keep)
  108. seurat_sp_small <- subset(seurat_sp, cells = sp_keep)
  110. # Check dimensions
  111. dim(seurat_std_small) # genes x 500 (or fewer)
  112. dim(seurat_sp_small)
  114. seurat_std_small <- NormalizeData(seurat_std_small)
  115. seurat_std_small <- FindVariableFeatures(seurat_std_small)
  116. seurat_std_small <- ScaleData(seurat_std_small)
  117. seurat_std_small <- RunPCA(seurat_std_small)
  118. seurat_std_small <- RunUMAP(seurat_std_small, dims = 1:20, verbose = FALSE, seed.use = 1337)
  119. seurat_std_small <- FindNeighbors(seurat_std_small)
  120. seurat_std_small <- FindClusters(seurat_std_small, resolution = 0.4)
  123. # metric functions --------------------------------------------------------
  124. # ===========================
  125. # Helpers (species-agnostic mito detection)
  126. # ===========================
  127. .mito_patterns_default <- c("^MT-", "^mt-", "^Mt-") # human, mouse, mixed
  128. .mito_regex <- function(patterns) paste0("(", paste(patterns, collapse = "|"), ")")
  130. .get_mito_features <- function(seu, mito_patterns = NULL) {
  131. if (is.null(mito_patterns)) mito_patterns <- .mito_patterns_default
  132. genes <- rownames(seu)
  133. if (is.null(genes)) {
  134. return(character(0))
  135. }
  136. genes[grepl(.mito_regex(mito_patterns), genes)]
  137. }
  139. .get_mito_index_sce <- function(sce, mito_patterns = NULL) {
  140. if (is.null(mito_patterns)) mito_patterns <- .mito_patterns_default
  141. rn <- rownames(sce)
  142. if (is.null(rn)) {
  143. return(rep(FALSE, nrow(sce)))
  144. }
  145. grepl(.mito_regex(mito_patterns), rn)
  146. }
  148. # ===========================
  149. # 1) Utility: safe add meta by barcode (unchanged)
  150. # ===========================
  151. add_meta <- function(seu, df, suffix = NULL) {
  152. stopifnot(all(rownames(df) %in% colnames(seu)))
  153. if (!is.null(suffix)) colnames(df) <- paste0(colnames(df), "_", suffix)
  154. df <- df[colnames(seu), , drop = FALSE]
  155. Seurat::AddMetaData(seu, df)
  156. }
  158. # ---- META-ONLY MERGE (donor -> target) (minor namespace tighten; no behavior change) ----
  159. merge_meta_only <- function(target, donor, cols = NULL, suffix = "_donor",
  160. normalizer = .normalize_cells, verbose = TRUE) {
  161. stopifnot(inherits(target, "Seurat"), inherits(donor, "Seurat"))
  163. # maps of original -> normalized IDs
  164. map <- function(obj) {
  165. data.frame(
  166. orig = colnames(obj),
  167. norm = normalizer(colnames(obj)),
  168. stringsAsFactors = FALSE
  169. )
  170. }
  171. mT <- map(target)
  172. mD <- map(donor)
  174. # if normalization creates duplicate IDs, fall back to exact IDs
  175. if (any(duplicated(mT$norm)) || any(duplicated(mD$norm))) {
  176. if (verbose) message("Duplicates after normalization; falling back to exact barcode match.")
  177. mT$norm <- mT$orig
  178. mD$norm <- mD$orig
  179. }
  181. # intersection and index vectors
  182. common <- intersect(mT$norm, mD$norm)
  183. if (!length(common)) stop("No overlapping cells between target and donor after alignment.")
  184. idxT <- match(common, mT$norm)
  185. idxD <- match(common, mD$norm)
  187. # donor metadata (optionally subset columns)
  188. donor_meta <- [email hidden]
  189. if (!is.null(cols)) {
  190. missing_cols <- setdiff(cols, colnames(donor_meta))
  191. if (length(missing_cols)) warning("Skipping missing donor columns: ", paste(missing_cols, collapse = ", "))
  192. cols <- intersect(cols, colnames(donor_meta))
  193. donor_meta <- donor_meta[, cols, drop = FALSE]
  194. }
  196. # align rows to target’s order; fill NA where target has non-overlapping cells
  197. aligned <- donor_meta[mD$orig[idxD], , drop = FALSE]
  198. rownames(aligned) <- mT$orig[idxT]
  199. out <- matrix(NA,
  200. nrow = ncol(target), ncol = ncol(aligned),
  201. dimnames = list(colnames(target), colnames(aligned))
  202. )
  203. out[rownames(aligned), ] <- as.matrix(aligned)
  204. out <- as.data.frame(out, check.names = FALSE)
  206. # handle name collisions
  207. collide <- intersect(colnames([email hidden]), colnames(out))
  208. if (length(collide)) colnames(out)[match(collide, colnames(out))] <- paste0(collide, suffix)
  210. # add to target meta (by barcode rownames)
  211. target <- Seurat::AddMetaData(target, out)
  212. if (verbose) {
  213. message("Merged ", ncol(out), " column(s) onto target for ", length(common), " overlapping cells.")
  214. }
  215. return(target)
  216. }
  218. # ===========================
  219. # 2) Baseline: Seurat thresholds (species-agnostic via features list)
  220. # ===========================
  221. baseline_seurat <- function(seu, min_genes = 200, max_genes = 10000, max_mito = 10, mito_patterns = NULL) {
  222. mito_features <- .get_mito_features(seu, mito_patterns)
  223. if (length(mito_features)) {
  224. seu[["percent.mt"]] <- Seurat::PercentageFeatureSet(seu, features = mito_features)
  225. } else {
  226. warning("No mitochondrial features matched; setting percent.mt to NA.")
  227. seu[["percent.mt"]] <- NA_real_
  228. }
  229. keep <- with(
  230. [email hidden],
  231. nFeature_RNA > min_genes & nFeature_RNA < max_genes &
  232. (is.na(percent.mt) | percent.mt < max_mito)
  233. )
  234. add_meta(seu, data.frame(keep_seurat = keep, row.names = colnames(seu)))
  235. }
  237. # ===========================
  238. # 3) Baseline: scater/scuttle outlier rules (species-agnostic)
  239. # ===========================
  240. baseline_scater <- function(seu, mito_patterns = NULL) {
  241. sce <- as.SingleCellExperiment(seu)
  242. is_mito <- .get_mito_index_sce(sce, mito_patterns)
  243. sce <- scuttle::addPerCellQCMetrics(sce, subsets = list(mito = is_mito))
  244. low_lib <- scuttle::isOutlier(SummarizedExperiment::colData(sce)$sum, nmads = 3, type = "lower", log = TRUE)
  245. low_feat <- scuttle::isOutlier(SummarizedExperiment::colData(sce)$detected, nmads = 3, type = "lower", log = TRUE)
  246. high_mito <- scuttle::isOutlier(SummarizedExperiment::colData(sce)$subsets_mito_percent, nmads = 3, type = "higher")
  247. keep <- !(low_lib | low_feat | high_mito)
  248. add_meta(seu, data.frame(
  249. keep_scater = keep,
  250. subsets_mito_percent = SummarizedExperiment::colData(sce)$subsets_mito_percent,
  251. row.names = colnames(seu)
  252. ))
  253. }
  255. # ===========================
  256. # 4) Baseline: miQC (robust + species-agnostic; defaults unchanged)
  257. # ===========================
  258. baseline_miQC <- function(seu, mito_patterns = NULL, debug = TRUE) {
  259. stopifnot(inherits(seu, "Seurat"))
  260. sce <- as.SingleCellExperiment(seu)
  261. # ensure counts exist
  262. if (is.null(SummarizedExperiment::assay(sce, "counts"))) {
  263. SummarizedExperiment::assay(sce, "counts") <- Seurat::GetAssayData(seu, layer = "counts")
  264. }
  266. # Prepare required metrics (detected & subsets_mito_percent) in a species-agnostic way
  267. is_mito <- .get_mito_index_sce(sce, mito_patterns)
  268. sce <- scater::addPerCellQC(sce, subsets = list(mito = rownames(sce)[is_mito])) # adds 'detected' & 'subsets_mito_percent'
  270. det <- as.numeric(SummarizedExperiment::colData(sce)$detected)
  271. mito_pct <- as.numeric(SummarizedExperiment::colData(sce)$subsets_mito_percent) # 0–100
  273. # Guard against non-finite/degenerate inputs to flexmix
  274. bad <- !is.finite(det) | !is.finite(mito_pct)
  275. if (any(bad)) {
  276. if (isTRUE(debug)) message(sprintf("[miQC] dropping %d cells with non-finite QC metrics before fit", sum(bad)))
  277. sce <- sce[, !bad, drop = FALSE]
  278. }
  279. mito_now <- as.numeric(SummarizedExperiment::colData(sce)$subsets_mito_percent)
  280. if (length(unique(mito_now)) <= 1L) {
  281. # Avoid zero-variance crashes; if still degenerate, skip miQC gracefully
  282. if (isTRUE(debug)) message("[miQC] degenerate mitochondrial metric; skipping model and keeping all cells.")
  283. seu$miQC_prob <- NA_real_
  284. seu$keep_miQC <- TRUE
  285. return(seu)
  286. }
  288. # Fit miQC (linear → spline → 1D)
  289. mmod <- try(miQC::mixtureModel(sce, model_type = "linear"), silent = TRUE)
  290. if (inherits(mmod, "try-error")) {
  291. mmod <- try(miQC::mixtureModel(sce, model_type = "spline"), silent = TRUE)
  292. }
  293. if (inherits(mmod, "try-error")) {
  294. mmod <- try(miQC::mixtureModel(sce, model_type = "one_dimensional"), silent = TRUE)
  295. }
  296. if (inherits(mmod, "try-error")) {
  297. warning("miQC::mixtureModel failed after all model types; keeping all cells.")
  298. seu$miQC_prob <- NA_real_
  299. seu$keep_miQC <- TRUE
  300. return(seu)
  301. }
  303. # Posterior & choose “compromised” component (higher mito at median detected)
  304. post_mat <- flexmix::posterior(mmod)
  305. coef_mat <- flexmix::parameters(mmod)
  306. det_med <- stats::median(SummarizedExperiment::colData(sce)$detected, na.rm = TRUE)
  307. pick_comp <- function(coefs, det_med) {
  308. if (is.null(dim(coefs))) {
  309. return(2L)
  310. } # one_dimensional
  311. if (all(c("(Intercept)", "detected") %in% rownames(coefs))) {
  312. pred <- coefs["(Intercept)", ] + det_med * coefs["detected", ]
  313. return(which.max(pred)) # higher predicted mito% = compromised
  314. } else {
  315. return(2L)
  316. }
  317. }
  318. comp_k <- pick_comp(coef_mat, det_med)
  320. post_comp <- post_mat[, comp_k, drop = TRUE]
  321. names(post_comp) <- if (!is.null(rownames(post_mat))) rownames(post_mat) else colnames(sce)
  323. # Align back to Seurat barcodes
  324. seu$miQC_prob <- post_comp[match(colnames(seu), names(post_comp))]
  326. # Keep criterion: keep cells with posterior < 0.75 (miQC default)
  327. seu$keep_miQC <- seu$miQC_prob < 0.75
  329. if (isTRUE(debug)) {
  330. dir.create("qc_outputs", showWarnings = FALSE, recursive = TRUE)
  331. df <- data.frame(
  332. barcode = colnames(sce),
  333. detected = SummarizedExperiment::colData(sce)$detected,
  334. subsets_mito_percent = SummarizedExperiment::colData(sce)$subsets_mito_percent,
  335. miQC_prob = seu$miQC_prob[match(colnames(sce), colnames(seu))],
  336. keep_miQC = seu$keep_miQC[match(colnames(sce), colnames(seu))],
  337. stringsAsFactors = FALSE
  338. )
  339. utils::write.csv(df, file = "qc_outputs/miqc_input_metrics.csv", row.names = FALSE)
  340. }
  341. return(seu)
  342. }
  345. # ===========================
  346. # 7) Baseline: DropletQC damaged-only (requires nuclear_fraction + umi)
  347. # ===========================
  349. baseline_dropletqc <- function(seu, seurat_sp) {
  350. ## 1) Grab counts for spliced/unspliced and compute per-cell totals
  351. spliced <- GetAssayData(seurat_sp, assay = "spliced", layer = "counts")
  352. unspliced <- GetAssayData(seurat_sp, assay = "unspliced", layer = "counts")
  354. # Sanity: same cells in same order
  355. stopifnot(identical(colnames(spliced), colnames(unspliced)))
  357. exon_sum <- Matrix::colSums(spliced) # exonic counts
  358. intron_sum <- Matrix::colSums(unspliced) # intronic counts
  360. # DropletQC nuclear fraction = intronic / (intronic + exonic)
  361. nf <- intron_sum / (intron_sum + exon_sum + 1e-12)
  362. umi <- intron_sum + exon_sum # UMI for nf/umi plane (do NOT include ambiguous here)
  364. ## 2) Add to Seurat metadata
  365. seurat_sp$nuclear_fraction <- as.numeric(nf[colnames(seurat_sp)])
  366. seurat_sp$umi_nf_plane <- as.numeric(umi[colnames(seurat_sp)])
  368. # 1) Build the data frame for identify_empty_drops
  369. nf_umi <- data.frame(
  370. nf = as.numeric(seurat_sp$nuclear_fraction),
  371. umi = as.numeric(seurat_sp$umi_nf_plane),
  372. row.names = colnames(seurat_sp)
  373. )
  375. # Optional: drop any rows with NA (NF/UMI)
  376. nf_umi <- nf_umi[complete.cases(nf_umi[, 1:2]), ]
  378. # 2) Run identify_empty_drops
  379. ed <- identify_empty_drops(nf_umi = nf_umi, include_plot = FALSE)
  381. # 3) Add cell_type for identify_damaged_cells using the standard object's clusters
  382. # Use .normalize_cells internally if names dont perfectly match, but here they should due to subset alignment
  383. ed$cell_type <- seu$seurat_clusters[match(.normalize_cells(rownames(ed)), .normalize_cells(colnames(seu)))]
  385. # Drop any cell types that mapped to NA before pushing into identify_damaged_cells, else EM will crash
  386. ed <- ed[!is.na(ed$cell_type), ]
  388. # 4) Run DropletQC’s damaged-cell classifier
  389. dc <- identify_damaged_cells(
  390. ed, # min NF separation between intact vs damaged (tune if needed)
  391. verbose = TRUE,
  392. output_plots = T
  393. )
  394. dc$plots
  395. table(dc[[1]]$cell_status)
  397. # 5) Save calls back to Seurat metadata
  398. calls <- dc[[1]] # first list element = annotated table
  399. seurat_sp$DropletQC_status <- calls$cell_status[match(colnames(seurat_sp), rownames(calls))]
  400. seurat_sp$DropletQC_status[is.na(seurat_sp$DropletQC_status)] <- "unknown_NA"
  402. # labels can be "cell", "empty_droplet", or "damaged_cell"
  403. kept_dq <- seurat_sp$DropletQC_status == "cell"
  404. kept_dq[is.na(kept_dq)] <- FALSE
  405. seurat_sp$keep_dropletqc <- kept_dq
  407. want <- c("nuclear_fraction", "umi_nf_plane", "DropletQC_status", "keep_dropletqc")
  408. seu_merged_meta <- merge_meta_only(seu, seurat_sp, cols = want, suffix = "")
  409. return(seu_merged_meta)
  410. }
  413. ## qc plot -----------------------------------------------------------------
  416. seurat_std_small <- baseline_seurat(seurat_std_small)
  417. table(seurat_std_small$keep_seurat)
  419. p <- DimPlot(seurat_std_small, group.by = "keep_seurat", cols = c("red", "gray"), pt.size = 0.5, alpha = 0.5) +
  420. ggtitle("Seurat") +
  421. NoLegend()
  422. ggsave("keep_seurat.png", p, width = 4, height = 4)
  424. seurat_std_small <- baseline_scater(seurat_std_small)
  425. table(seurat_std_small$keep_scater)
  427. p <- DimPlot(seurat_std_small, group.by = "keep_scater", cols = c("red", "gray"), pt.size = 0.5, alpha = 0.5) +
  428. ggtitle("scater") +
  429. NoLegend()
  430. ggsave("keep_scater.png", p, width = 4, height = 4)
  433. seurat_std_small <- baseline_miQC(seurat_std_small)
  434. if (table(seurat_std_small$keep_miQC)[1] > table(seurat_std_small$keep_miQC)[2]) {
  435. seurat_std_small$keep_miQC <- !seurat_std_small$keep_miQC
  436. }
  437. table(seurat_std_small$keep_miQC)
  439. p <- DimPlot(seurat_std_small, group.by = "keep_miQC", cols = c("red", "gray"), pt.size = 0.5, alpha = 0.5) +
  440. ggtitle("miQC") +
  441. NoLegend()
  442. p
  443. ggsave("keep_miqc.png", p, width = 4, height = 4)
  444. table(seurat_std_small$keep_miQC)
  447. seurat_std_small <- baseline_dropletqc(seurat_std_small, seurat_sp_small)
  448. table(seurat_std_small$keep_dropletqc)
  449. p <- DimPlot(seurat_std_small, group.by = "keep_dropletqc", cols = c("red", "gray"), pt.size = 0.5, alpha = 0.5) +
  450. ggtitle("DropletQC") +
  451. NoLegend()
  452. p
  453. ggsave("keep_dq.png", p, width = 4, height = 4)
  456. # run scqc ----------------------------------------------------------------
  459. library(scQCenrich)
  460. res <- run_qc_pipeline(
  461. obj = seurat_std_small, # normal Seurat for baseline QC
  462. species = organism,
  463. assay = "RNA",
  464. method = "gmm",
  465. # external splicing sources:
  466. spliced_obj = seurat_sp_small,
  467. spliced_assay = "spliced",
  468. spliced_layer = "counts",
  469. unspliced_obj = seurat_sp_small,
  470. unspliced_assay = "unspliced",
  471. unspliced_layer = "counts",
  472. report_html = T,
  473. report_file = "qc_outputs/qc_report.html",
  474. debug = TRUE,
  475. annot_method = "marker_score",
  476. tissue = organ,
  477. marker_method = "findmarkers",
  478. doublets = "none",
  479. enrichment_plots = T
  480. # qc_strength = "strict"
  481. )
  482. if (!is.null(res$report)) browseURL(res$report)
  484. seu <- res$obj_all
  485. seu$keep_enrich <- seu$qc_status != "remove"
  486. seurat_std_small$keep_enrich <- seu$keep_enrich[match(colnames(seurat_std_small), colnames(seu))]
  489. p <- DimPlot(seu, group.by = "keep_enrich", cols = c("red", "gray"), pt.size = 0.5, alpha = 0.5) +
  490. ggtitle("scQCenrich") +
  491. NoLegend()
  492. p
  493. ggsave("keep_enrich.png", p, width = 4, height = 4)
  498. # ---- scQCenrich WITHOUT splice data -------------------------
  501. cat("\n--- Running scQCenrich without splice data (MALAT1 proxy) ---\n")
  502. met_nosplice <- calcQCmetrics(
  503. obj = seurat_std_small,
  504. species = organism,
  505. assay = "RNA",
  506. add_to_meta = FALSE
  507. # spliced_obj intentionally omitted -> MALAT1 proxy activates
  508. )
  509. cat(
  510. "intronic_frac all-NA (confirming no splice used):",
  511. all(is.na(met_nosplice$intronic_frac)), "\n"
  512. )
  513. cat("MALAT1_frac range:", round(range(met_nosplice$MALAT1_frac, na.rm = TRUE), 4), "\n")
  515. res_nosplice <- flagLowQuality(
  516. metrics = met_nosplice,
  517. method = "gmm",
  518. qc_strength = "auto",
  519. rescue_mode = "lenient"
  520. )
  522. # Align barcodes back to seurat_std_small (metrics rownames == colnames(seu))
  523. seurat_std_small$keep_enrich_nosplice <- res_nosplice$qc_status[
  524. match(colnames(seurat_std_small), rownames(res_nosplice))
  525. ] != "remove"
  527. cat("kept (no splice):", sum(seurat_std_small$keep_enrich_nosplice, na.rm = TRUE), "\n")
  528. cat("removed (no splice):", sum(!seurat_std_small$keep_enrich_nosplice, na.rm = TRUE), "\n")
  530. p_nosplice <- DimPlot(
  531. seurat_std_small,
  532. group.by = "keep_enrich_nosplice",
  533. cols = c("red", "gray"), pt.size = 0.5, alpha = 0.5
  534. ) +
  535. ggtitle("scQCenrich (no splice)") +
  536. NoLegend()
  537. p_nosplice
  538. ggsave("keep_enrich_nosplice.png", p_nosplice, width = 4, height = 4)

benchmarking.R at commit fa00b9c, under other · at the source

Overview

Authors: Yuanyuan Liu1,2, Cheng Yang3, Chenghui Wang4, Mingwang Zhang5, Kai Luo6, Lihua Wu6, Xiufeng Xie6
ORCID iDs: Yuanyuan Liu
  1. The Seventh Medical Center of Chinese PLA General Hospital, Beijing, China
  2. PLA Medical School, Beijing, China
  3. Reahealth Inc, Beijing, China
  4. Bioinformatics Center of AMMS, Beijing, China
  5. Department of Dermatology, Southwest Hospital, Army Medical University, Chongqing, China
  6. Biological Therapy Center, Seventh Medical Center of Chinese PLA General Hospital, Beijing, China
Institutions: Chinese PLA General Hospital (China); Chinese People's Liberation Army (China); Army Medical University (China); Southwest Hospital (China)
Journal: Communications biology, volume 9, issue 1, article 864
Dates: received 13 January 2026; accepted 21 May 2026; published online 24 June 2026
Type: Research article · Language: English
License: CC BY-NC-ND
Identifiers: DOI 10.1038/s42003-026-10382-x · PMID 42342867 · PMCID PMC13315592 · OpenAlex W7165783598
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: genetics / omics (modality), human (organism), mouse (organism), cellular / molecular (subfield)
Methods: Statistics, Smoothing, state filtering, decompositions, Connectivity, Preprocessing
Keywords: Data processing, RNA sequencing
MeSH: Sequence Analysis, RNA*, Single-Cell Analysis*, Animals, Humans, Mice, Quality Control, Single-Cell Gene Expression Analysis (* major topic)
Topic: Single-cell and spatial transcriptomics (Molecular Biology, Biochemistry, Genetics and Molecular Biology), according to OpenAlex
Citations: cited by 1 paper (Europe PMC); 26 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.

Repositories

Its files are read in the Code ↔ Paper reader above, with 13 matches between paragraphs and lines of code.

lemonlyy755/scQCenrich

License: other
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: fa00b9c0126bdaa5ba36a6b90a28c734d5397a6b, 7 May 2026
Languages: R (28)
Size: 75 files, 28 scripts
Software Heritage: not archived
Found in: “Code availability”
Holds: README, license file, CITATION.cff, environment (DESCRIPTION), tests, continuous integration, documentation, 3 notebooks
Tools: Seurat (17 files), ggplot2 (7 files), tidyverse (5 files), clusterProfiler (3 files), SingleCellExperiment (3 files), patchwork (2 files), circlize (1 file), ComplexHeatmap (1 file), lme4 (1 file), reshape2 (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
31 files
At the source: github.com/lemonlyy755/scQCenrich

Zenodo 20050798

License: CC-BY-4.0
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Size: 1 file
Software Heritage: not checked
Found in: the references
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: Seurat (17 files), ggplot2 (7 files), tidyverse (5 files), clusterProfiler (3 files), SingleCellExperiment (3 files), patchwork (2 files), circlize (1 file), ComplexHeatmap (1 file), lme4 (1 file), reshape2 (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)
  • 27 September 2026: the link answers (HTTP 200)
31 files
At the source:

Code availability statement

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

Read it in the paper: doi.org/10.1038/s42003-026-10382-x.

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;
  • 56 scripts, each with its path and the digest of its content;
  • 13 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 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:

Read it in the paper: doi.org/10.1038/s42003-026-10382-x.

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, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 7 authors, 2 keywords, 7 MeSH terms, 22 references.

Cite

This paper

Liu, Y., Yang, C., Wang, C., Zhang, M., Luo, K., Wu, L., & Xie, X. (2026). ScQCenrich enables multi-metric quality control for single-cell RNA sequencing. Communications biology, 9(1), 864. https://doi.org/10.1038/s42003-026-10382-x

BibTeX

@article{liu2026scqcenrich,
author = {Liu, Yuanyuan and Yang, Cheng and Wang, Chenghui and Zhang, Mingwang and Luo, Kai and Wu, Lihua and Xie, Xiufeng},
title = {{ScQCenrich enables multi-metric quality control for single-cell RNA sequencing}},
journal = {Communications biology},
year = {2026},
month = jun,
volume = {9},
number = {1},
pages = {864},
publisher = {Nature Publishing Group},
issn = {2399-3642},
doi = {10.1038/s42003-026-10382-x},
url = {https://doi.org/10.1038/s42003-026-10382-x},
pmid = {42342867},
pmcid = {PMC13315592}
}

RIS

TY - JOUR
AU - Liu, Yuanyuan
AU - Yang, Cheng
AU - Wang, Chenghui
AU - Zhang, Mingwang
AU - Luo, Kai
AU - Wu, Lihua
AU - Xie, Xiufeng
TI - ScQCenrich enables multi-metric quality control for single-cell RNA sequencing
T2 - Communications biology
J2 - Commun Biol
PY - 2026
DA - 2026/06/24
VL - 9
IS - 1
SP - 864
SN - 2399-3642
PB - Nature Publishing Group
DO - 10.1038/s42003-026-10382-x
UR - https://doi.org/10.1038/s42003-026-10382-x
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s42003-026-10382-x",
"type": "article-journal",
"title": "ScQCenrich enables multi-metric quality control for single-cell RNA sequencing",
"container-title": "Communications biology",
"author": [
{
"family": "Liu",
"given": "Yuanyuan"
},
{
"family": "Yang",
"given": "Cheng"
},
{
"family": "Wang",
"given": "Chenghui"
},
{
"family": "Zhang",
"given": "Mingwang"
},
{
"family": "Luo",
"given": "Kai"
},
{
"family": "Wu",
"given": "Lihua"
},
{
"family": "Xie",
"given": "Xiufeng"
}
],
"container-title-short": "Commun Biol",
"volume": "9",
"issue": "1",
"page": "864",
"DOI": "10.1038/s42003-026-10382-x",
"PMID": "42342867",
"PMCID": "PMC13315592",
"ISSN": "2399-3642",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s42003-026-10382-x",
"language": "en",
"issued": {
"date-parts": [
[
2026,
6,
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.crmeth.2026.101337 [code]
Comparative analysis of nuclei isolation methods for brain single-nucleus RNA sequencing.
Journal: Cell reports methods
In common: circlize, ComplexHeatmap, Seurat, 3 other tools, genetics / omics, mouse, cellular / molecular, 7 references
[2] 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, clusterProfiler, 7 other tools, genetics / omics, cellular / molecular, 2 references
[3] doi:10.1038/s41467-026-76232-w [code]
Th17 effector cytokines induce shared and distinct microglial and endothelial cell responses in a mouse model for post-streptococcal encephalitis.
Journal: Nature communications
In common: SingleCellExperiment, circlize, clusterProfiler, 6 other tools, genetics / omics, mouse, cellular / molecular, 1 reference
[4] doi:10.1038/s41467-026-69944-6 [code]
Multi-modal dissection of cell-type specific TDP-43 pathology in the motor cortex.
Journal: Nature communications
In common: circlize, clusterProfiler, ComplexHeatmap, 6 other tools, genetics / omics, 1 reference
[5] doi:10.1002/imt2.70163 [code]
Spatial multi-omics unveils sphingolipid metabolic reprogramming within the retinal pathological niche.
Journal: iMeta
In common: SingleCellExperiment, circlize, clusterProfiler, 6 other tools, genetics / omics, mouse, cellular / molecular
[6] doi:10.1111/adb.70179 [code]
Transcriptional Response to Chronic Long-Access Fentanyl Self-Administration in Rat Habenula and Amygdala.
Journal: Addiction biology
In common: SingleCellExperiment, circlize, clusterProfiler, 6 other tools, genetics / omics, cellular / molecular
[7] doi:10.1038/s41586-026-10214-2 [code]
Multidimensional profiling of heterogeneity in supratentorial ependymomas.
Journal: Nature
In common: SingleCellExperiment, circlize, clusterProfiler, 6 other tools, genetics / omics, mouse
[8] doi:10.1038/s41593-026-02384-z [code]
cGAS-mediated type I IFN signaling contributes to disease progression in drug-refractory epilepsy.
Journal: Nature neuroscience
In common: SingleCellExperiment, circlize, clusterProfiler, 6 other tools, mouse, cellular / molecular
[9] doi:10.1016/j.celrep.2026.117500 [code]
Spatio-molecular gene expression reflects dorsal anterior cingulate cortex structure and function in the human brain.
Journal: Cell reports
In common: SingleCellExperiment, circlize, ComplexHeatmap, 5 other tools, genetics / omics, cellular / molecular, 1 reference
[10] doi:10.1038/s42003-026-10034-0 [code]
Region- and cell type-specific changes in gene expression in the cerebellum after classical fear conditioning.
Journal: Communications biology
In common: SingleCellExperiment, circlize, clusterProfiler, 5 other tools, genetics / omics, mouse, cellular / molecular, 1 reference

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.