biopipen 0.34.6__py3-none-any.whl → 0.34.26__py3-none-any.whl

biopipen/scripts/scrna/PseudoBulkDEG.R

@@ -1,6 +1,7 @@
 library(rlang)
 library(dplyr)
 library(plotthis)
+library(Seurat)
 library(biopipen.utils)
 
 sobjfile <- {{in.sobjfile | r}}
@@ -8,7 +9,9 @@ outdir <- {{out.outdir | r}}
 joboutdir <- {{job.outdir | r}}
 each <- {{envs.each | r}}
 subset <- {{envs.subset | r}}
+ncores <- {{envs.ncores | r}}
 mutaters <- {{envs.mutaters | r}}
+cache <- {{ envs.cache | r }}
 aggregate_by <- {{envs.aggregate_by | r}}
 layer <- {{envs.layer | r}}
 assay <- {{envs.assay | r}}
@@ -35,6 +38,7 @@ overlaps <- {{ envs.overlaps | r }}
 cases <- {{envs.cases | r}}
 
 aggregate_by <- unique(c(aggregate_by, group_by, paired_by, each))
+if (isTRUE(cache)) { cache <- joboutdir }
 
 log <- get_logger()
 reporter <- get_reporter()
@@ -74,10 +78,12 @@ defaults <- list(
     ident_1 = ident_1,
     ident_2 = ident_2,
     dbs = dbs,
+    ncores = ncores,
     sigmarkers = sigmarkers,
     enrich_style = enrich_style,
     paired_by = paired_by,
     tool = tool,
+    cache = cache,
     allmarker_plots_defaults = allmarker_plots_defaults,
     allmarker_plots = allmarker_plots,
     allenrich_plots_defaults = allenrich_plots_defaults,
@@ -131,12 +137,14 @@ expand_each <- function(name, case) {
 
         if (length(cases) == 0 && name == "DEG Analysis") {
             name <- case$each
+        } else {
+            name <- paste0(name, " (", case$each, ")")
         }
 
         case$aggregate_by <- unique(c(case$aggregate_by, case$group_by, case$paired_by, case$each))
 
         for (each in eachs) {
-            newname <- paste0(case$each, "::", each)
+            newname <- paste0(name, "::", each)
             newcase <- case
 
             newcase$original_case <- name
@@ -179,6 +187,7 @@ expand_each <- function(name, case) {
         if (length(case$overlaps) > 0 || length(case$allmarker_plots) > 0 || length(case$allenrich_plots) > 0) {
             ovcase <- case
 
+            ovcase$allexprs <- list()
             ovcase$markers <- list()
             ovcase$allmarker_plots <- lapply(
                 ovcase$allmarker_plots,
@@ -212,7 +221,52 @@ process_markers <- function(markers, info, case) {
     # markers <- markers %>%
     #     mutate(gene = as.character(gene)) %>%
     #     arrange(p_val_adj, desc(abs(avg_log2FC)))
+
+    empty <- if (case$enrich_style == "enrichr") {
+        data.frame(
+            Database = character(0),
+            Term = character(0),
+            Overlap = character(0),
+            P.value = numeric(0),
+            Adjusted.P.value = numeric(0),
+            Odds.Ratio = numeric(0),
+            Combined.Score = numeric(0),
+            Genes = character(0),
+            Rank = numeric(0)
+        )
+    } else {  # clusterProfiler
+        data.frame(
+            ID = character(0),
+            Description = character(0),
+            GeneRatio = character(0),
+            BgRatio = character(0),
+            Count = integer(0),
+            pvalue = numeric(0),
+            p.adjust = numeric(0),
+            qvalue = numeric(0),
+            geneID = character(0),
+            Database = character(0)
+        )
+    }
+    if (is.null(markers) || nrow(markers) == 0) {
+        if (case$error) {
+            stop("Error: No markers found in case '", info$name, "'.")
+        } else {
+            log$warn("! Warning: No markers found in case '", info$name, "'.")
+            reporter$add2(
+                list(
+                    name = "Warning",
+                    contents = list(list(kind = "error", content = "No markers found.", kind_ = "warning"))),
+                hs = c(info$section, info$name),
+                hs2 = "DEG Analysis",
+                ui = "tabs"
+            )
+            return(empty)
+        }
+    }
     markers$gene <- as.character(markers$gene)
+    markers$p_val_adj <- as.numeric(markers$p_val_adj)
+    markers$log2FC <- as.numeric(markers$log2FC)
     markers <- markers[order(markers$p_val_adj, -abs(markers$log2FC)), ]
 
     # Save markers
@@ -287,7 +341,7 @@ process_markers <- function(markers, info, case) {
             stop("Error: Not enough significant DEGs with '", case$sigmarkers, "' in case '", info$name, "' found (< 5) for enrichment analysis.")
         } else {
             message <- paste0("Not enough significant DEGs with '", case$sigmarkers, "' found (< 5) for enrichment analysis.")
-            log$warn("  ! Error: {message}")
+            log$warn("! Error: {message}")
             reporter$add2(
                 list(
                     name = "Warning",
@@ -345,7 +399,7 @@ process_markers <- function(markers, info, case) {
             if (case$error) {
                 stop("Error: ", e$message)
             } else {
-                log$warn("  ! Error: {e$message}")
+                log$warn("! Error: {e$message}")
                 reporter$add2(
                     list(
                         name = "Warning",
@@ -478,6 +532,7 @@ process_overlaps <- function(markers, ovcases, casename, groupname) {
 
 run_case <- function(name) {
     case <- cases[[name]]
+    log$info("----------------------------------------")
     log$info("Case: {name} ...")
 
     case <- extract_vars(
@@ -485,18 +540,21 @@ run_case <- function(name) {
         "dbs", "sigmarkers", "allmarker_plots", "allenrich_plots", "marker_plots", "enrich_plots",
         "overlaps", "original_case", "markers", "enriches", "each_name", "each", "enrich_style",
         "aggregate_by", "subset", "layer", "assay", "group_by", "ident_1", "ident_2", "original_subset",
-        "paired_by", "tool", "error",
+        "paired_by", "tool", "error", "ncores", "cache", "allexprs",
         allow_nonexisting = TRUE
     )
 
     if (!is.null(markers) || !is.null(enriches)) {
-        if (!is.null(markers)) {  # It is the overlap/allmarker case
-            log$info("- Summarizing DEGs in subcases (by each: {each}) ...")
+        if (!is.null(markers) && length(markers) > 0) {
+            log$info("Summarizing DEGs in subcases (by each: {each}) ...")
             # handle the overlaps / allmarkers analysis here
             if (!is.data.frame(markers)) {
                 each_levels <- names(markers)
                 markers <- do_call(rbind, lapply(each_levels, function(x) {
                     markers_df <- markers[[x]]
+                    if (is.null(markers_df) || nrow(markers_df) == 0) {
+                        return(NULL)
+                    }
                     if (nrow(markers_df) > 0) {
                         markers_df[[each]] <- x
                     } else {
@@ -508,17 +566,17 @@ run_case <- function(name) {
             }
             # gene, p_val, avg_log2FC, pct.1, pct.2, p_val_adj, diff_pct, <each>
 
+            if (!is.data.frame(allexprs)) {
+                meta <- do_call(rbind, lapply(allexprs, attr, "meta"))
+                allexprs <- do_call(cbind, allexprs)
+            } else {
+                meta <- attr(allexprs, "meta")
+            }
+
             if (length(allmarker_plots) > 0) {
-                log$info("- Visualizing all DEGs together ...")
-                exprs <- AggregateExpressionPseudobulk(
-                    srtobj, aggregate_by = aggregate_by, layer = layer, assay = assay,
-                    subset = original_subset, log = log
-                )
-                attr(markers, "object") <- AggregateExpressionPseudobulk(
-                    srtobj, aggregate_by = aggregate_by, layer = layer, assay = assay,
-                    subset = original_subset, log = log
-                )
-                attr(markers, "meta") <- attr(exprs, "meta")
+                log$info("Visualizing all DEGs together ...")
+                attr(markers, "object") <- allexprs
+                attr(markers, "meta") <- meta
                 attr(markers, "group_by") <- each
                 attr(markers, "paired_by") <- paired_by
                 attr(markers, "ident_1") <- NULL
@@ -527,18 +585,21 @@ run_case <- function(name) {
             }
 
             if (length(overlaps) > 0) {
-                log$info("- Visualizing overlaps between subcases ...")
+                log$info("Visualizing overlaps between subcases ...")
                 process_overlaps(markers, overlaps, name, each)
             }
 
         }
 
-        if (!is.null(enriches)) {
-            log$info("- Summarizing enrichments in subcases (by each: {each}) ...")
+        if (!is.null(enriches) && length(enriches) > 0) {
+            log$info("Summarizing enrichments in subcases (by each: {each}) ...")
             if (!is.data.frame(enriches)) {
                 each_levels <- names(enriches)
                 enriches <- do_call(rbind, lapply(each_levels, function(x) {
                     enrich_df <- enriches[[x]]
+                    if (is.null(enrich_df) || nrow(enrich_df) == 0) {
+                        return(NULL)
+                    }
                     if (nrow(enrich_df) > 0) {
                         enrich_df[[each]] <- x
                     } else {
@@ -546,11 +607,13 @@ run_case <- function(name) {
                     }
                     enrich_df
                 }))
-                enriches[[each]] <- factor(enriches[[each]], levels = each_levels)
+                if (!is.null(enriches) && nrow(enriches) > 0) {
+                    enriches[[each]] <- factor(enriches[[each]], levels = each_levels)
+                }
             }
 
-            if (length(allenrich_plots) > 0) {
-                log$info("- Visualizing all enrichments together ...")
+            if (length(allenrich_plots) > 0 && !is.null(enriches) && nrow(enriches) > 0) {
+                log$info("Visualizing all enrichments together ...")
                 process_allenriches(enriches, allenrich_plots, name, each)
             }
         }
@@ -558,16 +621,36 @@ run_case <- function(name) {
         return(invisible())
     }
 
+    info <- case_info(name, outdir, create = TRUE)
     exprs <- AggregateExpressionPseudobulk(
         srtobj, aggregate_by = aggregate_by, layer = layer, assay = assay,
         subset = subset, log = log
     )
-    markers <- RunDEGAnalysis(
-        exprs, group_by = group_by, ident_1 = ident_1, ident_2 = ident_2,
-        paired_by = paired_by, tool = tool, log = log
+    markers <- tryCatch(
+        {
+            RunDEGAnalysis(
+                exprs, group_by = group_by, ident_1 = ident_1, ident_2 = ident_2,
+                paired_by = paired_by, tool = tool, log = log, ncores = ncores,
+                cache = cache
+            )
+        }, error = function(e) {
+            if (error) {
+                stop("Error: ", e$message)
+            } else {
+                log$warn("! Error: {e$message}")
+                reporter$add2(
+                    list(
+                        name = "Warning",
+                        contents = list(list(kind = "error", content = e$message, kind_ = "warning"))),
+                    hs = c(info$section, info$name),
+                    hs2 = "DEG Analysis",
+                    ui = "tabs"
+                )
+                return(invisible())
+            }
+        }
     )
 
-    info <- case_info(name, outdir, create = TRUE)
     enrich <- process_markers(markers, info = info, case = list(
         dbs = dbs,
         sigmarkers = sigmarkers,
@@ -579,9 +662,12 @@ run_case <- function(name) {
     ))
 
     if (!is.null(original_case) && !is.null(cases[[original_case]])) {
-        markers[[each_name]] <- each
+        if (!is.null(markers)) {
+            markers[[each_name]] <- each
+        }
         cases[[original_case]]$markers[[each]] <<- markers
         cases[[original_case]]$enriches[[each]] <<- enrich
+        cases[[original_case]]$allexprs[[each]] <<- exprs
     }
 
     invisible()

biopipen/scripts/scrna/ScFGSEA.R

@@ -10,6 +10,7 @@ mutaters <- {{envs.mutaters | r}}  # nolint
 group_by <- {{envs.group_by | default: envs["group-by"] | default: None | r}}  # nolint
 ident_1 <- {{envs.ident_1 | default: envs["ident-1"] | default: None | r}}  # nolint
 ident_2 <- {{envs.ident_2 | default: envs["ident-2"] | default: None | r}}  # nolint
+assay <- {{envs.assay | r}}  # nolint
 each <- {{envs.each | r}}  # nolint
 subset <- {{envs.subset | r}}  # nolint
 gmtfile <- {{envs.gmtfile | r}}  # nolint
@@ -33,9 +34,6 @@ alleach_plots <- lapply(alleach_plots, function(x) {
 
 log$info("Reading Seurat object ...")
 srtobj <- read_obj(srtfile)
-if (!"Identity" %in% colnames(srtobj@meta.data)) {
-    srtobj@meta.data$Identity <- Idents(srtobj)
-}
 
 if (!is.null(mutaters) && length(mutaters) > 0) {
     log$info("Mutating metadata columns ...")
@@ -46,6 +44,7 @@ defaults <- list(
     group_by = group_by,
     ident_1 = ident_1,
     ident_2 = ident_2,
+    assay = assay,
     each = each,
     subset = subset,
     gmtfile = gmtfile,
@@ -63,7 +62,7 @@ defaults <- list(
 expand_each <- function(name, case) {
     outcases <- list()
 
-    case$group_by <- case$group_by %||% "Identity"
+    case$group_by <- case$group_by %||% GetIdentityColumn(srtobj)
 
     if (is.null(case$each) || is.na(case$each) || nchar(case$each) == 0 || isFALSE(each)) {
         if (length(case$alleach_plots) > 0) {
@@ -82,11 +81,13 @@ expand_each <- function(name, case) {
         }
 
         if (length(cases) == 0 && name == "GSEA") {
-            name <- case$each
+            prefix <- case$each
+        } else {
+            prefix <- paste0(name, " (", case$each, ")")
         }
 
         for (each in eachs) {
-            newname <- paste0(case$each, "::", each)
+            newname <- paste0(prefix, "::", each)
             newcase <- case
 
             newcase$original_case <- paste0(name, " (all ", case$each,")")
@@ -142,6 +143,11 @@ do_case <- function(name) {
 
     if (!is.null(case$gseas)) {
 
+        if (length(case$gseas) == 0) {
+            log$warn("  No GSEA results found for case {name}. Skipping.")
+            return(invisible(NULL))
+        }
+
         each_levels <- names(case$gseas)
         gseas <- do_call(rbind, lapply(each_levels, function(x) {
             gsea_df <- case$gseas[[x]]
@@ -226,7 +232,7 @@ do_case <- function(name) {
         case$ident_2 <- "Other"
         allclasses[allclasses != case$ident_1] <- "Other"
     }
-    exprs <- GetAssayData(sobj, layer = "data")
+    exprs <- GetAssayData(sobj, layer = "data", assay = case$assay)
 
     # get preranks
     log$info("  Getting preranks...")
@@ -240,25 +246,16 @@ do_case <- function(name) {
         quote = FALSE
     )
     if (all(is.na(ranks))) {
-        if (length(allclasses) < 100) {
-            log$warn("  Ignoring this case because all gene ranks are NA and there are <100 cells.")
-            reporter$add2(
-                list(
-                    kind = "error",
-                    content = paste0("Not enough cells (n = ", length(allclasses), ") to run fgsea.")
-                ),
-                hs = c(info$section, info$name)
-            )
-            return(NULL)
-        } else {
-            stop(paste0(
-                "All gene ranks are NA (# cells = ",
-                length(allclasses),
-                "). ",
-                "It's probably due to high missing rate in the data. ",
-                "You may want to try a different `envs$method` for pre-ranking."
-            ))
-        }
+        log$warn("  All gene ranks are NA. It's probably due to high missing rate in the data.")
+        log$warn("  Case ignored, you may also try a different ranking method.")
+        reporter$add2(
+            list(
+                kind = "error",
+                content = "All gene ranks are NA. It's probably due to high missing rate in the data."
+            ),
+            hs = c(info$section, info$name)
+        )
+        return(invisible(NULL))
     }
 
     # run fgsea

biopipen/scripts/scrna/ScVelo.py

@@ -7,13 +7,21 @@ from diot import Diot  # type: ignore[import]
 import scanpy as sc
 import scvelo as scv
 import numpy as np
+import matplotlib
+matplotlib.use('Agg')
 import matplotlib.pyplot as plt
-from biopipen.utils.misc import logger
+from biopipen.utils.misc import logger, require_package
 from biopipen.scripts.scrna.seurat_anndata_conversion import (
     convert_seurat_to_anndata,
     convert_anndata_to_seurat,
 )
 
+require_package("scvelo", ">=0.3.3")
+from biopipen.scripts.scrna import scvelo_paga  # noqa: F401
+
+warnings.simplefilter("ignore", category=UserWarning)
+warnings.simplefilter("ignore", category=FutureWarning)
+warnings.simplefilter("ignore", category=DeprecationWarning)
 
 
 def SCVELO(
@@ -45,10 +53,6 @@ def SCVELO(
     dpi=100,
     fileprefix="",
 ):
-    warnings.simplefilter("ignore", category=UserWarning)
-    warnings.simplefilter("ignore", category=FutureWarning)
-    warnings.simplefilter("ignore", category=DeprecationWarning)
-
     os.chdir(os.path.expanduser(dirpath))
     if linear_reduction is None:
         sc.pp.pca(adata, n_comps=n_pcs)
@@ -526,18 +530,26 @@ calculate_velocity_genes: bool = {{envs.calculate_velocity_genes | repr}}  # pyr
 top_n: int = {{envs.top_n | repr}}  # pyright: ignore  # noqa: E999
 rscript: str = {{envs.rscript | repr}}  # pyright: ignore  # noqa: E999
 
-if group_by is None:
-    raise ValueError("The 'envs.group_by' parameter must be specified.")
 
 if sobjfile.endswith(".h5ad"):
     h5ad_file = Path(sobjfile)
 else:
     h5ad_file = Path(outfile).with_suffix(".input.h5ad")
     logger.info("Converting Seurat object to AnnData (h5ad) format...")
-    convert_seurat_to_anndata(
+    seurat_ident_col = convert_seurat_to_anndata(
         input_file=sobjfile,
         output_file=h5ad_file,
         rscript=rscript,
+        return_ident_col=not group_by,
+    )
+    group_by = group_by or seurat_ident_col
+
+if group_by is None:
+    group_by = "seurat_clusters"
+    logger.warning(
+        "`envs.group_by` is not provided. "
+        "Using 'seurat_clusters' as the default groupby column. "
+        "It is recommended to provide the `envs.group_by` parameter."
     )
 
 logger.info(f"Reading AnnData (h5ad) file ...")

biopipen/scripts/scrna/SeuratClusterStats-clustree.R

@@ -16,7 +16,7 @@ if (
             if (startsWith(key, "FindClusters") && length(srtobj@commands[[key]]$resolution) > 1) {
                 pref <- substring(key, 14)
                 if (pref == "") {
-                    pref <- "seurat_clusters"
+                    pref <- biopipen.utils::GetIdentityColumn(srtobj)
                 }
 
                 clustrees[[pref]] <- list(prefix = pref)

biopipen/scripts/scrna/SeuratClusterStats-features.R

@@ -107,7 +107,12 @@ do_one_features <- function(name) {
         caching$restore()
     } else {
         case$features <- .get_features(features, case$object)
-        p <- do_call(gglogger::register(FeatureStatPlot), case)
+        p <- tryCatch({
+            do_call(gglogger::register(FeatureStatPlot), case)
+        }, error = function(e) {
+            if (save_code) { stop(e) }
+            do_call(FeatureStatPlot, case)
+        })
         save_plot(p, info$prefix, devpars, formats = c("png", more_formats))
         if (save_code) {
             save_plotcode(p, info$prefix,

biopipen/scripts/scrna/SeuratClustering.R

@@ -1,4 +1,5 @@
 
+library(rlang)
 library(Seurat)
 library(biopipen.utils)
 
@@ -11,13 +12,16 @@ RunPCAArgs <- {{envs.RunPCA | r: todot="-"}}
 FindNeighborsArgs <- {{envs.FindNeighbors | r: todot="-"}}
 FindClustersArgs <- {{envs.FindClusters | r: todot="-"}}
 RunUMAPArgs <- {{envs.RunUMAP | r: todot="-"}}
+ident <- {{envs.ident | r }}
 cache <- {{envs.cache | r}}
 ncores <- {{envs.ncores | r}}
 
+FindClustersArgs$cluster.name <- FindClustersArgs$cluster.name %||% ident %||% "seurat_clusters"
+
 log <- get_logger()
 
 # options(str = strOptions(vec.len = 5, digits.d = 5))
-options(future.globals.maxSize = 80000 * 1024^2)
+options(future.globals.maxSize = Inf)
 plan(strategy = "multicore", workers = ncores)
 
 log$info("Reading Seurat object ...")

biopipen/scripts/scrna/SeuratMap2Ref.R

@@ -25,7 +25,7 @@ plots = {{envs.plots | r}}
 log <- get_logger()
 reporter <- get_reporter()
 
-options(future.globals.maxSize = 8 * 1024 ^ 4)
+options(future.globals.maxSize = Inf)
 options(future.rng.onMisuse="ignore")
 options(Seurat.object.assay.version = "v5")
 
@@ -43,7 +43,6 @@ if (isTRUE(cache)) {
     cache = joboutdir
 }
 if (is.null(split_by)) {
-    options(future.globals.maxSize = 8 * 1024 ^ 4)
     future::plan(strategy = "multicore", workers = ncores)
 }
 

biopipen/scripts/scrna/SeuratPreparing.R

@@ -17,7 +17,7 @@ reporter <- get_reporter()
 
 set.seed(8525)
 # 8TB
-options(future.globals.maxSize = 8 * 1024 ^ 4)
+options(future.globals.maxSize = Inf)
 options(future.rng.onMisuse="ignore")
 options(Seurat.object.assay.version = "v5")
 plan(strategy = "multicore", workers = envs$ncores)
@@ -38,19 +38,27 @@ reporter$add(
     h1 = "Filters and QC"
 )
 
-metadata <- read.table(
-    metafile,
-    header = TRUE,
-    row.names = NULL,
-    sep = "\t",
-    check.names = FALSE
-)
+metadata <- tryCatch({
+    log$debug("Trying to read Seurat object from metafile ...")
+    read_obj(metafile)
+}, error = function(e) {
+    log$debug("Failed to read Seurat object from metafile: {e$message}")
+    log$debug("Reading metafile as a table (sample info) ...")
+    read.table(
+        metafile,
+        header = TRUE,
+        row.names = NULL,
+        sep = "\t",
+        check.names = FALSE
+    )
+})
+is_seurat <- inherits(metadata, "Seurat")
 
-meta_cols = colnames(metadata)
+meta_cols <- if (is_seurat) colnames(metadata@meta.data) else colnames(metadata)
 if (!"Sample" %in% meta_cols) {
-    stop("Error: Column `Sample` is not found in metafile.")
+    stop("Error: Column `Sample` is not found in ", ifelse(is_seurat, "Seurat object's meta.data.", "metafile."))
 }
-if (!"RNAData" %in% meta_cols) {
+if (!"RNAData" %in% meta_cols && !is_seurat) {
     stop("Error: Column `RNAData` is not found in metafile.")
 }
 

biopipen/scripts/scrna/SeuratSubClustering.R

@@ -17,7 +17,7 @@ FindNeighborsArgs <- {{envs.FindNeighbors | r: todot = "-"}}
 FindClustersArgs <- {{envs.FindClusters | r: todot = "-"}}
 cases <- {{envs.cases | r}}
 
-options(future.globals.maxSize = 80000 * 1024^2)
+options(future.globals.maxSize = Inf)
 plan(strategy = "multicore", workers = ncores)
 
 log <- get_logger()

biopipen/scripts/scrna/Slingshot.R

@@ -16,16 +16,14 @@ align_start <- {{envs.align_start | r}}
 seed <- {{envs.seed | r}}
 
 set.seed(seed)
-if (is.null(group_by)) {
-    stop("envs.group_by is required")
-}
 
 log <- get_logger()
 
 log$info("Reading Seurat object ...")
 srt <- read_obj(sobjfile)
+group_by <- group_by %||% biopipen.utils::GetIdentityColumn(srt)
 
-if (!group_by %in% colnames(srt@meta.data)) {
+if (is.null(group_by) || !group_by %in% colnames(srt@meta.data)) {
     stop(paste("Grouping column", group_by, "not found in the Seurat object"))
 }
 

biopipen/scripts/scrna/TopExpressingGenes.R

@@ -25,9 +25,6 @@ reporter <- get_reporter()
 
 log$info("Reading Seurat object ...")
 srtobj <- read_obj(srtfile)
-if (!"Identity" %in% colnames(srtobj@meta.data)) {
-    srtobj@meta.data$Identity <- Idents(srtobj)
-}
 assay <- DefaultAssay(srtobj)
 
 if (!is.null(mutaters) && length(mutaters) > 0) {
@@ -171,7 +168,7 @@ run_case <- function(name) {
     } else {
         subobj <- srtobj
     }
-    case$group_by <- case$group_by %||% "Identity"
+    case$group_by <- case$group_by %||% GetIdentityColumn(srtobj)
     if (is.null(case$ident)) {
         case$ident <- as.character(unique(subobj@meta.data[[case$group_by]]))
     }