Skip to contents

Integration workflow

Source: R/integration.R
RunIntegration.Rd

Integrate single-cell data with one or more methods. For ChromatinAssay, the workflow uses TFIDF + SVD/LSI; Uncorrected is supported directly, Harmony5 is redirected to Harmony, and Seurat/RPCA are not supported.

Usage

RunIntegration(
  srt_merge = NULL,
  batch,
  append = TRUE,
  srt_list = NULL,
  assay = NULL,
  integration_methods = c("Uncorrected", "Seurat", "CCA", "RPCA", "scVI", "PeakVI",
    "PoissonVI", "WNN", "MultiMAP", "GLUE", "scVI5", "MNN", "fastMNN", "fastMNN5",
    "Harmony", "Harmony5", "Scanorama", "BBKNN", "CSS", "Coralysis", "LIGER", "Conos",
    "ComBat"),
  compute_lisi = FALSE,
  lisi_label_colnames = NULL,
  lisi_reduction = NULL,
  lisi_dims = NULL,
  lisi_prefix = NULL,
  lisi_tool_name = NULL,
  lisi_perplexity = 30,
  lisi_tol = 1e-05,
  lisi_max_iter = 50,
  lisi_knn_algorithm = c("auto", "brute_force", "clustered"),
  lisi_cores = NULL,
  lisi_max_dense_bytes = Inf,
  compute_metrics = FALSE,
  metrics_batch_col = NULL,
  metrics_celltype_col = NULL,
  metrics_reduction = NULL,
  metrics_cluster_col = NULL,
  metrics_tool_name = NULL,
  metrics_k_graph = 15,
  do_normalization = NULL,
  normalization_method = "LogNormalize",
  do_HVF_finding = TRUE,
  HVF_source = "separate",
  HVF_method = "vst",
  nHVF = 2000,
  HVF_min_intersection = 1,
  HVF = NULL,
  do_scaling = TRUE,
  vars_to_regress = NULL,
  regression_model = "linear",
  scale_within_batch = FALSE,
  linear_reduction = "pca",
  linear_reduction_dims = 50,
  linear_reduction_dims_use = NULL,
  linear_reduction_params = list(),
  force_linear_reduction = FALSE,
  nonlinear_reduction = "umap",
  nonlinear_reduction_dims = c(2, 3),
  nonlinear_reduction_params = list(),
  force_nonlinear_reduction = TRUE,
  neighbor_metric = "euclidean",
  neighbor_k = 20L,
  cluster_algorithm = "louvain",
  cluster_resolution = 0.6,
  seed = 11,
  verbose = TRUE,
  integration_method = NULL,
  ...
)

Arguments

srt_merge

A merged `Seurat` object that includes the batch information.

batch

Batch variable name.

append

Append integrated results to srt_merge.

srt_list

A list of Seurat objects to be checked and preprocessed.

assay

Assay to use. NULL uses the default assay.

integration_methods

Method(s) to run. Multiple methods require append = TRUE and are applied sequentially. For ChromatinAssay, prefer "Uncorrected" or "Harmony5".

compute_lisi, lisi_label_colnames, lisi_reduction, lisi_dims, lisi_prefix, lisi_tool_name, lisi_perplexity, lisi_tol, lisi_max_iter, lisi_knn_algorithm, lisi_cores, lisi_max_dense_bytes

LISI scores. lisi_label_colnames = NULL uses batch when it is a single metadata column; lisi_reduction = NULL uses DefaultReduction().

compute_metrics, metrics_batch_col, metrics_celltype_col, metrics_reduction, metrics_cluster_col, metrics_tool_name, metrics_k_graph

Integration summary metrics on the selected reduction.

do_normalization

Whether data normalization should be performed.

normalization_method

The normalization method to be used. Possible values are "LogNormalize", "SCT", "TFIDF", and "scran".

do_HVF_finding, HVF_method, nHVF, HVF

Highly variable features. HVF_method is "vst", "mvp", "disp", or "scran".

HVF_source

The source of highly variable features. Possible values are "global" and "separate".

HVF_min_intersection

The feature needs to be present in batches for a minimum number of times in order to be considered as highly variable.

do_scaling

Force scaling via ScaleData.

vars_to_regress

A vector of variable names to include as additional regression variables.

regression_model

"linear", "poisson", or "negativebinomial".

scale_within_batch

Scale within each batch. Only used by "Uncorrected", "Seurat", "MNN", "Harmony", "BBKNN", "CSS", "ComBat".

linear_reduction, linear_reduction_dims, linear_reduction_dims_use, linear_reduction_params, force_linear_reduction

Linear reduction ("pca", "svd", "ica", "nmf", "mds", "glmpca"). linear_reduction_dims_use = NULL uses estimated dimensions unless the method specifies a different default; explicit values select computed components.

nonlinear_reduction

Nonlinear reduction ("umap", "umap-naive", "tsne", "dm", "phate", "pacmap", "trimap", "largevis", "fr").

nonlinear_reduction_dims

Number(s) of nonlinear dimensions to compute.

nonlinear_reduction_params

Shared parameters, optionally overridden by lists named after individual nonlinear reduction methods.

force_nonlinear_reduction

Recompute nonlinear reductions even when stored results exist.

neighbor_metric, neighbor_k

Neighbor graph ("euclidean", "cosine", "manhattan", "hamming").

cluster_algorithm, cluster_resolution

Clustering ("louvain", "slm", "leiden"). Larger cluster_resolution yields fewer clusters.

seed

Random seed.

verbose

Whether to print the message. Default is TRUE.

integration_method

Deprecated alias of integration_methods.

...

Arguments passed to other functions.

Value

A Seurat object. For ChromatinAssay, names follow the ATAC convention (*lsi, *UMAP2D, cluster aliases, ATAC_default_*).

See also

Seurat_integrate, scVI_integrate, MultiMAP_integrate, WNN_integrate, GLUE_integrate, MNN_integrate, fastMNN_integrate, Harmony_integrate, Scanorama_integrate, BBKNN_integrate, CSS_integrate, Coralysis_integrate, LIGER_integrate, Conos_integrate, ComBat_integrate, RunIntegrationBenchmark, IntegrationBenchmarkPlot

Examples

data(panc8_sub)
panc8_sub <- RunIntegration(
  panc8_sub,
  batch = "tech",
  integration_methods = "Harmony",
  nHVF = 500,
  linear_reduction_dims = 20,
  linear_reduction_dims_use = 1:10,
  nonlinear_reduction_dims = 2,
  compute_lisi = TRUE,
  lisi_label_colnames = "tech",
  lisi_perplexity = 10
)
#> ◌ [2026-11-12 05:39:14] Run integration workflow...
#> ℹ [2026-11-12 05:39:14] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:39:14] Checking a list of <Seurat>...
#> ! [2026-11-12 05:39:15] Data 1/5 of the `srt_list` is "unknown"
#> ℹ [2026-11-12 05:39:15] Perform `NormalizeData()` with `normalization.method = 'LogNormalize'` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:15] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ! [2026-11-12 05:39:15] Data 2/5 of the `srt_list` is "unknown"
#> ℹ [2026-11-12 05:39:15] Perform `NormalizeData()` with `normalization.method = 'LogNormalize'` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:15] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ! [2026-11-12 05:39:15] Data 3/5 of the `srt_list` is "unknown"
#> ℹ [2026-11-12 05:39:15] Perform `NormalizeData()` with `normalization.method = 'LogNormalize'` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:15] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ! [2026-11-12 05:39:15] Data 4/5 of the `srt_list` is "unknown"
#> ℹ [2026-11-12 05:39:15] Perform `NormalizeData()` with `normalization.method = 'LogNormalize'` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:15] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ! [2026-11-12 05:39:15] Data 5/5 of the `srt_list` is "unknown"
#> ℹ [2026-11-12 05:39:15] Perform `NormalizeData()` with `normalization.method = 'LogNormalize'` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:15] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:16] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:39:16] Number of available HVF: 500
#> ℹ [2026-11-12 05:39:16] Finished check
#> ℹ [2026-11-12 05:39:18] Perform `Seurat::ScaleData()`
#> ℹ [2026-11-12 05:39:19] Perform linear dimension reduction("pca")
#> ℹ [2026-11-12 05:39:19] Perform Harmony integration
#> ℹ [2026-11-12 05:39:19] Using "Harmonypca" (1:10) as input
#> ℹ [2026-11-12 05:39:19] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:39:20] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:39:20] Reorder clusters...
#> ℹ [2026-11-12 05:39:20] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:39:20] Perform umap nonlinear dimension reduction using Harmony (1:10)
#> ℹ [2026-11-12 05:39:27] Perform umap nonlinear dimension reduction using Harmonypca (1:10)
#> ℹ [2026-11-12 05:39:34] Compute LISI scores from reduction "HarmonypcaUMAP2D"
#> ℹ [2026-11-12 05:39:34] Compute LISI scores from reduction "HarmonyUMAP2D"
#> ✔ [2026-11-12 05:39:34] Stored LISI scores in metadata: "HarmonypcaUMAP2D_tech_LISI" and "HarmonyUMAP2D_tech_LISI"
#> ✔ [2026-11-12 05:39:35] Harmony integration completed
CellDimPlot(
  panc8_sub,
  group.by = c("tech", "celltype"),
  reduction = "HarmonyUMAP2D"
)


IntegrationBenchmarkPlot(panc8_sub, plot_type = "box")


panc8_sub <- RunIntegration(
  panc8_sub,
  batch = "tech",
  integration_methods = "LIGER"
)
#> ◌ [2026-11-12 05:39:35] Run integration workflow...
#> ℹ [2026-11-12 05:39:36] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:39:37] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:39:37] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:39:37] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:37] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:39:37] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:38] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:39:38] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:38] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:39:38] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:38] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:39:38] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:39:38] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:39:38] Number of available HVF: 2000
#> ℹ [2026-11-12 05:39:38] Finished check
#> ℹ [2026-11-12 05:39:39] Prepare rliger layer "ligerScaleData" ...
#> ℹ [2026-11-12 05:39:39] Perform LIGER integration
#> ℹ [2026-11-12 05:39:44] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:39:44] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:39:44] Reorder clusters...
#> ℹ [2026-11-12 05:39:45] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:39:45] Perform umap nonlinear dimension reduction using LIGER (1:20)
#> ℹ [2026-11-12 05:39:52] Perform umap nonlinear dimension reduction using LIGER (1:20)
#> ℹ [2026-11-12 05:39:59] Perform umap nonlinear dimension reduction using LIGER (1:20)
#> ✔ [2026-11-12 05:40:07] LIGER integration completed
panc8_sub <- RunIntegration(
  panc8_sub,
  batch = "tech",
  integration_methods = "Harmony",
  compute_lisi = TRUE,
  lisi_label_colnames = "tech"
)
#> ◌ [2026-11-12 05:40:07] Run integration workflow...
#> ℹ [2026-11-12 05:40:08] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:40:09] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:40:09] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:09] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:09] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:09] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:09] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:09] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:10] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:10] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:10] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:10] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:10] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:40:10] Number of available HVF: 2000
#> ℹ [2026-11-12 05:40:10] Finished check
#> ℹ [2026-11-12 05:40:10] Perform `Seurat::ScaleData()`
#> ℹ [2026-11-12 05:40:10] Perform linear dimension reduction("pca")
#> ℹ [2026-11-12 05:40:11] Perform Harmony integration
#> ℹ [2026-11-12 05:40:11] Using "Harmonypca" (1:20) as input
#> ℹ [2026-11-12 05:40:11] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:40:12] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:40:12] Reorder clusters...
#> ℹ [2026-11-12 05:40:12] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:40:12] Perform umap nonlinear dimension reduction using Harmony (1:20)
#> ℹ [2026-11-12 05:40:19] Perform umap nonlinear dimension reduction using Harmony (1:20)
#> ℹ [2026-11-12 05:40:26] Perform umap nonlinear dimension reduction using Harmonypca (1:20)
#> ℹ [2026-11-12 05:40:33] Compute LISI scores from reduction "HarmonypcaUMAP2D"
#> ℹ [2026-11-12 05:40:33] Compute LISI scores from reduction "HarmonyUMAP2D"
#> ✔ [2026-11-12 05:40:33] Stored LISI scores in metadata: "HarmonypcaUMAP2D_tech_LISI" and "HarmonyUMAP2D_tech_LISI"
#> ✔ [2026-11-12 05:40:34] Harmony integration completed
IntegrationBenchmarkPlot(panc8_sub, plot_type = "box")


data("pbmcmultiome_sub", package = "scop")
pbmcmultiome_sub$batch <- rep(c("batch1", "batch2"), length.out = ncol(pbmcmultiome_sub))
pbmcmultiome_sub <- RunIntegration(
  pbmcmultiome_sub,
  batch = "batch",
  assay = "peaks",
  integration_methods = "Harmony5",
  normalization_method = "TFIDF"
)
#> ◌ [2026-11-12 05:40:34] Run integration workflow...
#> ! [2026-11-12 05:40:34] `integration_method = 'Harmony5'` is not compatible with <ChromatinAssay> in current Seurat v5 workflow. Automatically switch to "Harmony"
#> ℹ [2026-11-12 05:40:34] Split `srt_merge` into `srt_list` by "batch"
#> ℹ [2026-11-12 05:40:34] Checking a list of <Seurat>...
#> ! [2026-11-12 05:40:35] Data 1/2 of the `srt_list` is "raw_counts"
#> ℹ [2026-11-12 05:40:35] Perform `RunTFIDF()` on 1/2 of `srt_list`...
#> ℹ [2026-11-12 05:40:35] Perform `FindTopFeatures()` on 1/2 of `srt_list`...
#> ! [2026-11-12 05:40:35] Data 2/2 of the `srt_list` is "raw_counts"
#> ℹ [2026-11-12 05:40:35] Perform `RunTFIDF()` on 2/2 of `srt_list`...
#> ℹ [2026-11-12 05:40:35] Perform `FindTopFeatures()` on 2/2 of `srt_list`...
#> ℹ [2026-11-12 05:40:35] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:40:35] Number of available HVF: 2000
#> ℹ [2026-11-12 05:40:35] Finished check
#> ℹ [2026-11-12 05:40:35] `normalization_method` is "TFIDF". Use lsi workflow...
#> ℹ [2026-11-12 05:40:35] Perform linear dimension reduction("svd")
#> Running SVD
#> Scaling cell embeddings
#> ℹ [2026-11-12 05:40:36] Perform Harmony integration
#> ℹ [2026-11-12 05:40:36] Using "Harmonysvd" (2:5) as input
#> ℹ [2026-11-12 05:40:36] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:40:36] Reorder clusters...
#> ℹ [2026-11-12 05:40:36] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:40:36] Perform umap nonlinear dimension reduction using Harmony (1:4)
#> ℹ [2026-11-12 05:40:42] Perform umap nonlinear dimension reduction using Harmony (1:4)
#> ✔ [2026-11-12 05:40:48] Harmony integration completed

integration_methods <- c(
  "Uncorrected", "Seurat", "CCA", "RPCA",
  "MNN", "fastMNN", "Harmony", "Harmony5",
  "Coralysis", "LIGER", "Conos", "ComBat"
)
p_list <- list()
for (method in integration_methods) {
  panc8_sub <- RunIntegration(
    panc8_sub,
    batch = "tech",
    integration_methods = method,
    linear_reduction_dims_use = 1:50,
    nonlinear_reduction = "umap",
    RunParallelDivisiveICP_params = list(L = 10L)
  )
  p_list[[method]] <- CellDimPlot(
    panc8_sub,
    group.by = c("tech", "celltype"),
    reduction = paste0(method, "UMAP2D"),
    xlab = "", ylab = "",
    title = method,
    legend.position = "none",
    theme_use = "theme_blank"
  )
}
#> ◌ [2026-11-12 05:40:48] Run integration workflow...
#> ℹ [2026-11-12 05:40:49] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:40:50] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:40:50] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:50] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:50] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:50] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:50] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:50] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:51] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:51] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:51] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:40:51] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:40:51] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:40:51] Number of available HVF: 2000
#> ℹ [2026-11-12 05:40:51] Finished check
#> ℹ [2026-11-12 05:40:51] Perform Uncorrected integration
#> ℹ [2026-11-12 05:40:51] Perform `Seurat::ScaleData()`
#> ℹ [2026-11-12 05:40:51] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:40:52] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:40:53] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:40:53] Reorder clusters...
#> ℹ [2026-11-12 05:40:53] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:40:53] Perform umap nonlinear dimension reduction using Uncorrectedpca (1:50)
#> ℹ [2026-11-12 05:40:59] Perform umap nonlinear dimension reduction using Uncorrectedpca (1:50)
#> ℹ [2026-11-12 05:41:06] Perform umap nonlinear dimension reduction using Uncorrectedpca (1:50)
#> ✔ [2026-11-12 05:41:13] Uncorrected integration completed
#> ◌ [2026-11-12 05:41:14] Run integration workflow...
#> ℹ [2026-11-12 05:41:14] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:41:15] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:41:15] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:41:15] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:41:16] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:41:16] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:41:16] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:41:16] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:41:16] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:41:16] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:41:16] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:41:16] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:41:16] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:41:17] Number of available HVF: 2000
#> ℹ [2026-11-12 05:41:17] Finished check
#> ℹ [2026-11-12 05:41:17] Perform FindIntegrationAnchors
#> ℹ [2026-11-12 05:41:40] Perform Seurat integration
#> ℹ [2026-11-12 05:41:59] Perform ScaleData on `srt_integrated`
#> ℹ [2026-11-12 05:42:00] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:42:01] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:42:01] Reorder clusters...
#> ℹ [2026-11-12 05:42:01] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:42:01] Perform umap nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:42:08] Perform umap nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:42:14] Perform umap nonlinear dimension reduction using Seuratpca (1:50)
#> ✔ [2026-11-12 05:42:25] Seurat integration completed
#> ◌ [2026-11-12 05:42:25] Run integration workflow...
#> Splitting ‘counts’, ‘data’ layers. Not splitting . If you would like to split other layers, set in `layers` argument.
#> ℹ [2026-11-12 05:42:26] Data appears already log-normalized. Skipping `Seurat::NormalizeData()`
#> ℹ [2026-11-12 05:42:26] Perform `Seurat::FindVariableFeatures()` per batch (`HVF_source = 'separate'`)
#> ℹ [2026-11-12 05:42:26] Number of available HVF: 2000
#> ℹ [2026-11-12 05:42:28] Perform `Seurat::ScaleData()` on split layers for Seurat v5 integration
#> ℹ [2026-11-12 05:42:28] Perform PCA on split layers before `Seurat::IntegrateLayers()`
#> ℹ [2026-11-12 05:42:29] Perform Seurat v5 integration with `CCAIntegration()`
#> ℹ [2026-11-12 05:42:36] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:42:36] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:42:36] Reorder clusters...
#> ℹ [2026-11-12 05:42:36] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:42:36] Perform umap nonlinear dimension reduction using CCA (1:50)
#> ℹ [2026-11-12 05:42:43] Perform umap nonlinear dimension reduction using CCA (1:50)
#> ℹ [2026-11-12 05:42:50] Perform umap nonlinear dimension reduction using pca (1:50)
#> ✔ [2026-11-12 05:42:59] CCA integration completed
#> ◌ [2026-11-12 05:42:59] Run integration workflow...
#> Splitting ‘counts’, ‘data’ layers. Not splitting . If you would like to split other layers, set in `layers` argument.
#> ℹ [2026-11-12 05:43:00] Data appears already log-normalized. Skipping `Seurat::NormalizeData()`
#> ℹ [2026-11-12 05:43:00] Perform `Seurat::FindVariableFeatures()` per batch (`HVF_source = 'separate'`)
#> ℹ [2026-11-12 05:43:01] Number of available HVF: 2000
#> ℹ [2026-11-12 05:43:02] Perform `Seurat::ScaleData()` on split layers for Seurat v5 integration
#> ℹ [2026-11-12 05:43:02] Perform PCA on split layers before `Seurat::IntegrateLayers()`
#> ℹ [2026-11-12 05:43:03] Perform Seurat v5 integration with `RPCAIntegration()`
#> ℹ [2026-11-12 05:43:15] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:43:15] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:43:15] Reorder clusters...
#> ℹ [2026-11-12 05:43:15] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:43:15] Perform umap nonlinear dimension reduction using RPCA (1:50)
#> ℹ [2026-11-12 05:43:22] Perform umap nonlinear dimension reduction using RPCA (1:50)
#> ℹ [2026-11-12 05:43:29] Perform umap nonlinear dimension reduction using pca (1:50)
#> ✔ [2026-11-12 05:43:38] RPCA integration completed
#> ◌ [2026-11-12 05:43:38] Run integration workflow...
#> ℹ [2026-11-12 05:43:40] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:43:40] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:43:40] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:43:40] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:43:40] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:43:40] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:43:40] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:43:41] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:43:41] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:43:41] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:43:41] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:43:41] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:43:41] Number of available HVF: 2000
#> ℹ [2026-11-12 05:43:41] Finished check
#> ℹ [2026-11-12 05:43:42] Perform MNN integration
#> ℹ [2026-11-12 05:44:27] Perform ScaleData
#> ℹ [2026-11-12 05:44:27] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:44:28] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:44:28] Reorder clusters...
#> ℹ [2026-11-12 05:44:28] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:44:28] Perform umap nonlinear dimension reduction using MNNpca (1:50)
#> ℹ [2026-11-12 05:44:36] Perform umap nonlinear dimension reduction using MNNpca (1:50)
#> ℹ [2026-11-12 05:44:43] Perform umap nonlinear dimension reduction using MNNpca (1:50)
#> ✔ [2026-11-12 05:44:53] MNN integration completed
#> ◌ [2026-11-12 05:44:53] Run integration workflow...
#> ℹ [2026-11-12 05:44:55] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:44:55] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:44:55] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:44:55] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:44:55] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:44:55] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:44:55] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:44:56] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:44:56] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:44:56] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:44:56] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:44:56] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:44:56] Number of available HVF: 2000
#> ℹ [2026-11-12 05:44:56] Finished check
#> ℹ [2026-11-12 05:44:56] Perform fastMNN integration
#> ℹ [2026-11-12 05:45:00] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:45:01] Reorder clusters...
#> ℹ [2026-11-12 05:45:01] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:45:01] Perform umap nonlinear dimension reduction using fastMNN (1:50)
#> ℹ [2026-11-12 05:45:08] Perform umap nonlinear dimension reduction using fastMNN (1:50)
#> ℹ [2026-11-12 05:45:15] Perform umap nonlinear dimension reduction using fastMNN (1:50)
#> ✔ [2026-11-12 05:45:24] fastMNN integration completed
#> ◌ [2026-11-12 05:45:24] Run integration workflow...
#> ℹ [2026-11-12 05:45:26] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:45:27] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:45:28] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:45:28] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:45:28] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:45:28] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:45:28] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:45:28] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:45:28] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:45:28] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:45:28] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:45:28] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:45:29] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:45:29] Number of available HVF: 2000
#> ℹ [2026-11-12 05:45:29] Finished check
#> ℹ [2026-11-12 05:45:29] Perform `Seurat::ScaleData()`
#> ℹ [2026-11-12 05:45:29] Perform linear dimension reduction("pca")
#> ℹ [2026-11-12 05:45:30] Perform Harmony integration
#> ℹ [2026-11-12 05:45:30] Using "Harmonypca" (1:50) as input
#> ℹ [2026-11-12 05:45:30] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:45:31] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:45:31] Reorder clusters...
#> ℹ [2026-11-12 05:45:31] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:45:31] Perform umap nonlinear dimension reduction using Harmony (1:50)
#> ℹ [2026-11-12 05:45:38] Perform umap nonlinear dimension reduction using Harmony (1:50)
#> ℹ [2026-11-12 05:45:45] Perform umap nonlinear dimension reduction using Harmonypca (1:50)
#> ✔ [2026-11-12 05:45:54] Harmony integration completed
#> ◌ [2026-11-12 05:45:54] Run integration workflow...
#> Splitting ‘counts’, ‘data’ layers. Not splitting . If you would like to split other layers, set in `layers` argument.
#> ℹ [2026-11-12 05:45:56] Data appears already log-normalized. Skipping `Seurat::NormalizeData()`
#> ℹ [2026-11-12 05:45:56] Perform `Seurat::FindVariableFeatures()` per batch (`HVF_source = 'separate'`)
#> ℹ [2026-11-12 05:45:57] Number of available HVF: 2000
#> ℹ [2026-11-12 05:45:58] Perform `Seurat::ScaleData()` on split layers for Seurat v5 integration
#> ℹ [2026-11-12 05:45:58] Perform PCA on split layers before `Seurat::IntegrateLayers()`
#> ℹ [2026-11-12 05:45:59] Perform Seurat v5 integration with `HarmonyIntegration()`
#> The `features` argument is ignored by `HarmonyIntegration`.
#> This message is displayed once per session.
#> ℹ [2026-11-12 05:46:00] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:46:01] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:46:01] Reorder clusters...
#> ℹ [2026-11-12 05:46:01] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:46:01] Perform umap nonlinear dimension reduction using Harmony5 (1:50)
#> ℹ [2026-11-12 05:46:08] Perform umap nonlinear dimension reduction using Harmony5 (1:50)
#> ℹ [2026-11-12 05:46:15] Perform umap nonlinear dimension reduction using pca (1:50)
#> ✔ [2026-11-12 05:46:24] Harmony5 integration completed
#> ◌ [2026-11-12 05:46:24] Run integration workflow...
#> ℹ [2026-11-12 05:46:26] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:46:27] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:46:27] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:46:27] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:46:28] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:46:28] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:46:28] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:46:28] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:46:28] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:46:28] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:46:28] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:46:28] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:46:28] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:46:29] Number of available HVF: 2000
#> ℹ [2026-11-12 05:46:29] Finished check
#> ℹ [2026-11-12 05:46:29] Perform Coralysis integration
#> Data in `logcounts` slot already of `dgCMatrix` class...
#> 2000/2000 features remain after filtering features with only zero values.
#> 
#> Building training set...
#> No. of cells is lower or equal than the no. of 'nclusters': 200 <= 500
#> All the cells will be included and the clustering step will be skipped!
#> Found more than one class "kcca" in cache; using the first, from namespace 'kernlab'
#> Also defined by ‘flexclust’
#> Found more than one class "kcca" in cache; using the first, from namespace 'kernlab'
#> Also defined by ‘flexclust’
#> No. of cells is lower or equal than the no. of 'nclusters': 200 <= 500
#> All the cells will be included and the clustering step will be skipped!
#> No. of cells is lower or equal than the no. of 'nclusters': 200 <= 500
#> All the cells will be included and the clustering step will be skipped!
#> No. of cells is lower or equal than the no. of 'nclusters': 200 <= 500
#> All the cells will be included and the clustering step will be skipped!
#> Training set successfully built.
#> 
#> Computing cluster seed.
#> 
#> Initializing divisive ICP clustering...
#> 
  |                                                                            
  |                                                                      |   0%
  |                                                                            
  |=======                                                               |  10%
  |                                                                            
  |==============                                                        |  20%
  |                                                                            
  |=====================                                                 |  30%
  |                                                                            
  |============================                                          |  40%
  |                                                                            
  |===================================                                   |  50%
  |                                                                            
  |==========================================                            |  60%
  |                                                                            
  |=================================================                     |  70%
  |                                                                            
  |========================================================              |  80%
  |                                                                            
  |===============================================================       |  90%
  |                                                                            
  |======================================================================| 100%
#> 
#> 
#> Divisive ICP clustering completed successfully.
#> 
#> Predicting cell cluster probabilities using ICP models...
#> Prediction of cell cluster probabilities completed successfully.
#> 
#> Multi-level integration completed successfully.
#> Divisive ICP: selecting ICP tables multiple of 4
#> ℹ [2026-11-12 05:47:58] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:47:58] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:47:58] Reorder clusters...
#> ℹ [2026-11-12 05:47:58] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:47:58] Perform umap nonlinear dimension reduction using Coralysis (1:50)
#> ℹ [2026-11-12 05:48:05] Perform umap nonlinear dimension reduction using Coralysis (1:50)
#> ℹ [2026-11-12 05:48:13] Perform umap nonlinear dimension reduction using Coralysis (1:50)
#> ✔ [2026-11-12 05:48:22] Coralysis integration completed
#> ◌ [2026-11-12 05:48:23] Run integration workflow...
#> ℹ [2026-11-12 05:48:24] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:48:26] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:48:26] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:48:26] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:48:26] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:48:26] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:48:26] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:48:26] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:48:26] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:48:27] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:48:27] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:48:27] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:48:27] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:48:27] Number of available HVF: 2000
#> ℹ [2026-11-12 05:48:27] Finished check
#> ℹ [2026-11-12 05:48:27] Prepare rliger layer "ligerScaleData" ...
#> ℹ [2026-11-12 05:48:27] Perform LIGER integration
#> ℹ [2026-11-12 05:48:33] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:48:33] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:48:33] Reorder clusters...
#> ℹ [2026-11-12 05:48:34] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:48:34] Perform umap nonlinear dimension reduction using LIGER (1:20)
#> ℹ [2026-11-12 05:48:41] Perform umap nonlinear dimension reduction using LIGER (1:20)
#> ℹ [2026-11-12 05:48:49] Perform umap nonlinear dimension reduction using LIGER (1:50)
#> ✔ [2026-11-12 05:48:58] LIGER integration completed
#> ◌ [2026-11-12 05:48:58] Run integration workflow...
#> ℹ [2026-11-12 05:49:00] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:49:01] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:49:01] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:01] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:02] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:02] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:02] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:02] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:02] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:02] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:02] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:02] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:03] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:49:03] Number of available HVF: 2000
#> ℹ [2026-11-12 05:49:03] Finished check
#> ℹ [2026-11-12 05:49:03] Perform ScaleData on the data 1 ...
#> ℹ [2026-11-12 05:49:03] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:49:03] Perform ScaleData on the data 2 ...
#> ℹ [2026-11-12 05:49:03] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:49:03] Perform ScaleData on the data 3 ...
#> ℹ [2026-11-12 05:49:03] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:49:04] Perform ScaleData on the data 4 ...
#> ℹ [2026-11-12 05:49:04] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:49:04] Perform ScaleData on the data 5 ...
#> ℹ [2026-11-12 05:49:04] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:49:05]  Perform Conos integration
#> ℹ [2026-11-12 05:49:05] Conos integration using pca (1:50) as input
#> ℹ [2026-11-12 05:49:10] Adjust neighbor k from 20 to 20 for small-sample clustering
#> ℹ [2026-11-12 05:49:10] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:49:10] Reorder clusters...
#> ℹ [2026-11-12 05:49:10] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:49:10] Perform umap nonlinear dimension reduction using Conos
#> ℹ [2026-11-12 05:49:18] Perform umap nonlinear dimension reduction using Conos
#> ℹ [2026-11-12 05:49:25] Perform umap nonlinear dimension reduction using ConosUMAP2D (1:50)
#> ✔ [2026-11-12 05:49:35] Conos integration completed
#> ◌ [2026-11-12 05:49:35] Run integration workflow...
#> ℹ [2026-11-12 05:49:37] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:49:38] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:49:39] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:39] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:39] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:39] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:39] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:39] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:39] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:39] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:40] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:49:40] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:49:40] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:49:40] Number of available HVF: 2000
#> ℹ [2026-11-12 05:49:40] Finished check
#> ℹ [2026-11-12 05:49:40] Perform ComBat integration
#> Found 138 genes with uniform expression within a single batch (all zeros); these will not be adjusted for batch.
#> ℹ [2026-11-12 05:49:41] Perform `Seurat::ScaleData()`
#> ℹ [2026-11-12 05:49:41] Perform linear dimension reduction ("pca")
#> ℹ [2026-11-12 05:49:42] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:49:42] Reorder clusters...
#> ℹ [2026-11-12 05:49:42] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:49:42] Perform umap nonlinear dimension reduction using ComBatpca (1:50)
#> ℹ [2026-11-12 05:49:50] Perform umap nonlinear dimension reduction using ComBatpca (1:50)
#> ℹ [2026-11-12 05:49:57] Perform umap nonlinear dimension reduction using ComBatpca (1:50)
#> ✔ [2026-11-12 05:50:07] ComBat integration completed

# Python-backed methods prepare a scVI/scvi-tools environment and run model
# training, so keep them separate from ordinary example checks.
if (FALSE) { # \dontrun{
panc8_sub <- RunIntegration(
  panc8_sub,
  batch = "tech",
  integration_methods = "scVI",
  train_params = list(max_epochs = 2L),
  nonlinear_reduction = "umap"
)
panc8_sub <- RunIntegration(
  panc8_sub,
  batch = "tech",
  integration_methods = "scVI5",
  IntegrateLayers_params = list(max_epochs = 2L),
  nonlinear_reduction = "umap"
)
} # }

nonlinear_reductions <- c(
  "umap", "tsne", "dm", "phate",
  "pacmap", "trimap", "largevis", "fr"
)
panc8_sub <- RunIntegration(
  panc8_sub,
  batch = "tech",
  integration_methods = "Seurat",
  linear_reduction_dims_use = 1:50,
  nonlinear_reduction = nonlinear_reductions,
  nonlinear_reduction_params = list(largevis = list(n_epochs = 1000L))
)
#> ◌ [2026-11-12 05:50:07] Run integration workflow...
#> ℹ [2026-11-12 05:50:10] Split `srt_merge` into `srt_list` by "tech"
#> ℹ [2026-11-12 05:50:11] Checking a list of <Seurat>...
#> ℹ [2026-11-12 05:50:11] Data 1/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:50:11] Perform `FindVariableFeatures()` on 1/5 of `srt_list`...
#> ℹ [2026-11-12 05:50:11] Data 2/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:50:11] Perform `FindVariableFeatures()` on 2/5 of `srt_list`...
#> ℹ [2026-11-12 05:50:11] Data 3/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:50:11] Perform `FindVariableFeatures()` on 3/5 of `srt_list`...
#> ℹ [2026-11-12 05:50:11] Data 4/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:50:12] Perform `FindVariableFeatures()` on 4/5 of `srt_list`...
#> ℹ [2026-11-12 05:50:12] Data 5/5 of the `srt_list` has been log-normalized
#> ℹ [2026-11-12 05:50:12] Perform `FindVariableFeatures()` on 5/5 of `srt_list`...
#> ℹ [2026-11-12 05:50:12] Use the separate HVF from `srt_list`
#> ℹ [2026-11-12 05:50:12] Number of available HVF: 2000
#> ℹ [2026-11-12 05:50:12] Finished check
#> ℹ [2026-11-12 05:50:12] Perform FindIntegrationAnchors
#> ℹ [2026-11-12 05:50:50] Perform Seurat integration
#> ℹ [2026-11-12 05:51:44] Perform ScaleData on `srt_integrated`
#> ℹ [2026-11-12 05:51:44] Perform "pca" linear dimension reduction
#> ℹ [2026-11-12 05:51:45] Perform `Seurat::FindClusters()` with "louvain"
#> ℹ [2026-11-12 05:51:46] Reorder clusters...
#> ℹ [2026-11-12 05:51:46] Skip `log1p()` because `layer = data` is not "counts"
#> ℹ [2026-11-12 05:51:46] Perform umap nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:51:53] Perform umap nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:52:00] Perform tsne nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:52:04] Perform tsne nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:52:10] Perform dm nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:53:45] Perform dm nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:53:47] Perform phate nonlinear dimension reduction using Seuratpca (1:50)
#> ◌ [2026-11-12 05:53:50] Preparing python environment...
#> ℹ [2026-11-12 05:53:50] Environment name: 'scprep_env' and python version: 3.10-1
#> ℹ [2026-11-12 05:53:50] Auto-detecting conda-compatible environment manager...
#> ℹ [2026-11-12 05:53:50] Using micromamba executable: '/home/runner/.cache/R/scop/micromamba/linux-64/bin/micromamba'
#> ℹ [2026-11-12 05:53:50] Creating Python environment with python "3.10-1" using micromamba...
#> ℹ [2026-11-12 05:53:50] Skipping conda Terms of Service acceptance for micromamba.
#> + /home/runner/.cache/R/scop/micromamba/linux-64/bin/micromamba 'create' '--yes' '--name' 'scprep_env' 'python=3.10-1' 'pip' 'setuptools<81' 'wheel' '--quiet' '-c' 'conda-forge'
#> ✔ [2026-11-12 05:53:53] Environment created successfully: '/home/runner/work/_temp/pkgdown-python/envs/scprep_env'
#> ℹ [2026-11-12 05:53:53] 13 requirements to check, 3 conda requirements and 10 pip requirements
#> ℹ [2026-11-12 05:53:53] Checking 3 conda requirements
#> ℹ [2026-11-12 05:53:53] Installing 3 packages into environment: 'scprep_env' using micromamba
#> ℹ [2026-11-12 05:53:54] Installing packages via micromamba...
#> ℹ [2026-11-12 05:53:54] Using channel: "conda-forge"
#> ℹ [2026-11-12 05:53:54] Installing 3 packages...
#> + /home/runner/.cache/R/scop/micromamba/linux-64/bin/micromamba 'install' '--yes' '--name' 'scprep_env' '-c' 'conda-forge' 'leidenalg=0.10.2' 'tbb=2022.2.0' 'python-igraph=0.11.9'
#> Using Cached Shard Index for conda-forge/linux-64                                                   ✔ Done
#> Using Cached Shard Index for conda-forge/noarch                                                     ✔ Done
#> Fetching and Parsing Packages' Shards                                                           ⧖ Starting
#> Fetching and Parsing Packages' Shards                                                     ✔ Done (0.2 sec)
#> 
#> Pinned packages:
#> 
#>   - python=3.10
#> 
#> Resolving Environment                                                                           ⧖ Starting
#> Resolving Environment                                                                     ✔ Done (0.4 sec)
#> 
#> Transaction
#> 
#>   Prefix: /home/runner/work/_temp/pkgdown-python/envs/scprep_env
#> 
#>   Updating specs:
#> 
#>    - leidenalg=0.10.2
#>    - tbb=2022.2.0
#>    - python-igraph=0.11.9
#> 
#> 
#>   Package           Version  Build                  Channel           Size
#> ────────────────────────────────────────────────────────────────────────────
#>   Install:
#> ────────────────────────────────────────────────────────────────────────────
#> 
#>   + arpack            3.9.1  nompi_hf03ea27_102     conda-forge     Cached
#>   + glpk                5.0  h445213a_0             conda-forge     Cached
#>   + gmp               6.3.0  hfd2156b_3             conda-forge     Cached
#>   + igraph          0.10.16  hfe3e89f_1             conda-forge     Cached
#>   + leidenalg        0.10.2  py310hea6c23e_1        conda-forge     Cached
#>   + libblas          3.11.0  12_h4a7cf45_openblas   conda-forge     Cached
#>   + libgfortran      16.2.0  h69a702a_7             conda-forge     Cached
#>   + libgfortran5     16.2.0  h6b99dfc_7             conda-forge     Cached
#>   + libhwloc         2.12.1  default_hafda6a7_1003  conda-forge     Cached
#>   + libiconv           1.18  h0cb94f2_3             conda-forge     Cached
#>   + liblapack        3.11.0  12_h47877c9_openblas   conda-forge     Cached
#>   + libleidenalg     0.11.1  h00ab1b0_0             conda-forge     Cached
#>   + libopenblas      0.3.34  pthreads_hf13c14d_2    conda-forge     Cached
#>   + libstdcxx        16.2.0  h934c35e_7             conda-forge     Cached
#>   + libstdcxx-ng     16.2.0  hdf11a46_7             conda-forge     Cached
#>   + libxml2          2.15.4  ha7a5771_1             conda-forge     Cached
#>   + libxml2-16       2.15.4  hbdfff7e_1             conda-forge     Cached
#>   + libxml2-devel    2.15.4  ha7a5771_1             conda-forge     Cached
#>   + python-igraph    0.11.9  py310hea6c23e_3        conda-forge     Cached
#>   + python_abi         3.10  9_cp310                conda-forge     Cached
#>   + tbb            2022.2.0  hb60516a_1             conda-forge     Cached
#>   + texttable         1.7.0  pyhd8ed1ab_1           conda-forge     Cached
#> 
#>   Summary:
#> 
#>   Install: 22 packages
#> 
#>   Total download: 0 B
#> 
#> ─────────────────────────────────────────────────warning  libmamba Security Warning: This transaction includes executing package scripts (pre/post-link/unlink) if present. These scripts can contain arbitrary code. Please ensure you trust the package sources.
#> ───────────────────────────
#> 
#> 
#> 
#> Transaction starting
#> Linking libstdcxx-16.2.0-h934c35e_7
#> Linking libiconv-1.18-h0cb94f2_3
#> Linking libgfortran5-16.2.0-h6b99dfc_7
#> Linking gmp-6.3.0-hfd2156b_3
#> Linking libstdcxx-ng-16.2.0-hdf11a46_7
#> Linking libxml2-16-2.15.4-hbdfff7e_1
#> Linking libgfortran-16.2.0-h69a702a_7
#> Linking glpk-5.0-h445213a_0
#> Linking libxml2-2.15.4-ha7a5771_1
#> Linking libopenblas-0.3.34-pthreads_hf13c14d_2
#> Linking libhwloc-2.12.1-default_hafda6a7_1003
#> Linking libxml2-devel-2.15.4-ha7a5771_1
#> Linking libblas-3.11.0-12_h4a7cf45_openblas
#> Linking tbb-2022.2.0-hb60516a_1
#> Linking liblapack-3.11.0-12_h47877c9_openblas
#> Linking arpack-3.9.1-nompi_hf03ea27_102
#> Linking igraph-0.10.16-hfe3e89f_1
#> Linking libleidenalg-0.11.1-h00ab1b0_0
#> Linking python_abi-3.10-9_cp310
#> Linking texttable-1.7.0-pyhd8ed1ab_1
#> Linking python-igraph-0.11.9-py310hea6c23e_3
#> Linking leidenalg-0.10.2-py310hea6c23e_1
#> 
#> Transaction finished
#> 
#> ✔ [2026-11-12 05:53:55] micromamba installation completed successfully
#> ℹ [2026-11-12 05:53:56] Checking 12 requirements using uv
#> ℹ [2026-11-12 05:53:56] Installing 12 packages into environment: 'scprep_env' using micromamba
#> ℹ [2026-11-12 05:53:56] Installing packages via uv...
#> + /home/runner/work/_temp/pkgdown-python/envs/scop_env/bin/uv 'pip' 'install' '--python' '/home/runner/work/_temp/pkgdown-python/envs/scprep_env/bin/python3.10' '--force-reinstall' 'setuptools<81' 'matplotlib==3.10.8' 'numba==0.66.0' 'llvmlite==0.48.0' 'numpy==1.26.4' 'packaging>=24.0' 'pandas==2.0.3' 'scikit-learn==1.7.0' 'scipy==1.15.3' 'phate==1.0.11' 'leidenalg==0.10.2' 'igraph==0.11.9'
#> Using Python 3.10.1 environment at: /home/runner/work/_temp/pkgdown-python/envs/scprep_env
#> Resolved 35 packages in 364ms
#> Downloading pandas (11.7MiB)
#> Downloading pygsp (1.8MiB)
#>  Downloaded pygsp
#>  Downloaded pandas
#> Prepared 35 packages in 293ms
#> Uninstalled 5 packages in 18ms
#> Installed 35 packages in 59ms
#>  + cloudpickle==3.1.2
#>  + contourpy==1.3.2
#>  + cycler==0.12.1
#>  + decorator==5.3.1
#>  + deprecated==1.3.1
#>  + fonttools==4.65.0
#>  + future==1.0.0
#>  + graphtools==2.1.0
#>  - igraph==0.11.9 (from file:///home/conda/feedstock_root/build_artifacts/python-igraph_1761728289959/work)
#>  + igraph==0.11.9
#>  + joblib==1.6.0
#>  + kiwisolver==1.5.1
#>  - leidenalg==0.10.2 (from file:///home/conda/feedstock_root/build_artifacts/leidenalg_1756310397572/work)
#>  + leidenalg==0.10.2
#>  + llvmlite==0.48.0
#>  + matplotlib==3.10.8
#>  + numba==0.66.0
#>  + numpy==1.26.4
#>  - packaging==26.3 (from file:///home/conda/feedstock_root/build_artifacts/bld/rattler-build_packaging_1785888127/work)
#>  + packaging==26.3
#>  + pandas==2.0.3
#>  + phate==1.0.11
#>  + pillow==12.3.0
#>  + pygsp==0.6.1
#>  + pyparsing==3.3.3
#>  + python-dateutil==3.0.0
#>  + pytz==2026.6
#>  + s-gd2==1.8.1
#>  + scikit-learn==1.7.0
#>  + scipy==1.15.3
#>  + scprep==1.2.3
#>  ~ setuptools==80.10.2
#>  + six==1.17.0
#>  + tasklogger==1.2.0
#>  - texttable==1.7.0 (from file:///home/conda/feedstock_root/build_artifacts/texttable_1734145585512/work)
#>  + texttable==1.7.2
#>  + threadpoolctl==3.8.0
#>  + tzdata==2026.6
#>  + wrapt==2.5.2
#> ✔ [2026-11-12 05:53:57] uv installation completed
#> + /home/runner/work/_temp/pkgdown-python/envs/scprep_env/bin/python3.10 -m pip check
#> No broken requirements found.
#> ✔ [2026-11-12 05:53:59] Python environment ready
#> ✔                       Until next loading, the environment will be cached
#> Environment manager config:
#>  Manager:       micromamba
#>  Executable:    /home/runner/.cache/R/scop/micromamba/linux-64/bin/micromamba
#>  Environment:   /home/runner/work/_temp/pkgdown-python/envs/scprep_env
#> Python config:
#>  python:        /home/runner/work/_temp/pkgdown-python/envs/scprep_env/bin/python3.10
#>  version:       Python 3.10.1
#> 
#> ℹ [2026-11-12 05:54:13] Perform phate nonlinear dimension reduction using Seuratpca (1:50)
#> ◌ [2026-11-12 05:54:16] Preparing python environment...
#> ℹ [2026-11-12 05:54:16] Environment name: 'scprep_env' and python version: 3.10-1
#> ℹ [2026-11-12 05:54:16] Auto-detecting conda-compatible environment manager...
#> ℹ [2026-11-12 05:54:16] Using micromamba executable: '/home/runner/.cache/R/scop/micromamba/linux-64/bin/micromamba'
#> ℹ [2026-11-12 05:54:16] Using existing environment: '/home/runner/work/_temp/pkgdown-python/envs/scprep_env'
#> ℹ [2026-11-12 05:54:16] 13 requirements to check, 3 conda requirements and 10 pip requirements
#> ℹ [2026-11-12 05:54:16] Checking 3 conda requirements
#> ℹ [2026-11-12 05:54:17] Checking 12 requirements using uv
#> + /home/runner/work/_temp/pkgdown-python/envs/scprep_env/bin/python3.10 -m pip check
#> No broken requirements found.
#> ✔ [2026-11-12 05:54:19] Python environment ready
#> ✔                       Until next loading, the environment will be cached
#> Environment manager config:
#>  Manager:       micromamba
#>  Executable:    /home/runner/.cache/R/scop/micromamba/linux-64/bin/micromamba
#>  Environment:   /home/runner/work/_temp/pkgdown-python/envs/scprep_env
#> Python config:
#>  python:        /home/runner/work/_temp/pkgdown-python/envs/scprep_env/bin/python3.10
#>  version:       Python 3.10.1
#> 
#> ℹ [2026-11-12 05:54:32] Perform pacmap nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:54:32] Perform pacmap nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:54:33] Perform trimap nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:54:33] Perform trimap nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:54:34] Perform largevis nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:54:44] Perform largevis nonlinear dimension reduction using Seuratpca (1:50)
#> ℹ [2026-11-12 05:54:53] Perform fr nonlinear dimension reduction using Seurat_SNN
#> ℹ [2026-11-12 05:54:54] Perform fr nonlinear dimension reduction using Seurat_SNN
#> ℹ [2026-11-12 05:54:57] Perform umap nonlinear dimension reduction using Seuratpca (1:50)
#> ✔ [2026-11-12 05:55:18] Seurat integration completed
for (nr in nonlinear_reductions) {
  print(
    CellDimPlot(
      panc8_sub,
      group.by = c("tech", "celltype"),
      reduction = paste0("Seurat", nr, "2D"),
      xlab = "", ylab = "", title = nr,
      legend.position = "none", theme_use = "theme_blank"
    )
  )
}