Version:	1.1.1
Title:	Adaptive Machine Learning-Powered, Context-Matching Tool for Single-Cell and Spatial Transcriptomics Annotation
Description:	Annotates single-cell and spatial-transcriptomic (ST) data using context-matching marker datasets. It creates a unified marker list (‘Markers_list') from multiple sources: built-in curated databases (’Cellmarker2', 'PanglaoDB', 'scIBD', 'TCellSI', 'PCTIT', 'PCTAM'), Seurat objects with cell labels, or user-provided Excel tables. SlimR first uses adaptive machine learning for parameter optimization, and then offers two automated annotation approaches: 'cluster-based' and 'per-cell'. Cluster-based annotation assigns one label per cluster, expression-based probability calculation, and AUC validation. Per-cell annotation assigns labels to individual cells using three scoring methods with adaptive thresholds and ratio-based confidence filtering, plus optional UMAP spatial smoothing, making it ideal for heterogeneous clusters and rare cell types. The package also supports semi-automated workflows with heatmaps, feature plots, and combined visualizations for manual annotation. For more details, see Kabacoff (2020, ISBN:9787115420572).
License:	MIT + file LICENSE
URL:	https://github.com/zhaoqing-wang/SlimR
BugReports:	https://github.com/zhaoqing-wang/SlimR/issues
Depends:	R (≥ 3.5)
Imports:	cowplot, dplyr, ggplot2, patchwork, pheatmap, readxl, scales, Seurat, tidyr, tools, tibble
Suggests:	crayon, RANN, testthat (≥ 3.0.0)
Encoding:	UTF-8
LazyData:	true
RoxygenNote:	7.3.3
Date:	2026-02-05
NeedsCompilation:	no
Packaged:	2026-02-05 15:49:32 UTC; Runaw
Author:	Zhaoqing Wang [aut, cre]
Maintainer:	Zhaoqing Wang <zhaoqingwang@mail.sdu.edu.cn>
Repository:	CRAN
Date/Publication:	2026-02-05 16:20:14 UTC

Apply UMAP-based spatial smoothing to scores

Description

Apply UMAP-based spatial smoothing to scores

Usage

.apply_umap_smoothing(
  seurat_obj,
  score_matrix,
  umap_reduction,
  k_neighbors,
  smoothing_weight,
  chunk_size,
  verbose
)

Compute AUCell-like rank-based scores

Description

Uses a ranking approach similar to AUCell: for each cell, genes are ranked by expression, and the score is based on where marker genes fall in that ranking. This method is robust to batch effects and technical variation.

Usage

.compute_aucell_scores(expr_matrix, marker_sets, top_percent = 0.05)

Details

Key improvement: Uses recovery curve area under curve (AUC) calculation rather than simple proportion, giving partial credit to markers ranked just outside the top threshold.

Compute weighted scores for per-cell annotation

Description

This function uses an improved weighting scheme that considers:

1. Expression level (log-normalized)

2. Detection rate (binary: above min_expression threshold)

3. Marker specificity (how unique is this marker to this cell type)

4. Expression variability (CV-based: more variable genes are more discriminative)

Usage

.compute_weighted_scores(expr_matrix, marker_sets, min_expression)

Cellmarker2 dataset

Description

A dataset containing marker genes for different cell types from Cellmarker2

Usage

Cellmarker2

Format

A data frame with 8 columns:

Details

This dataset is used to filter and create a standardized marker list. The dataset can be filtered based on species, tissue class, tissue type, cancer type, and cell type to generate a list of marker genes for specific cell types.

Source

http://117.50.127.228/CellMarker/

See Also

Other Section_0_Database: Cellmarker2_raw, Cellmarker2_table, Markers_list_PCTAM, Markers_list_PCTIT, Markers_list_TCellSI, Markers_list_scIBD, PanglaoDB, PanglaoDB_raw, PanglaoDB_table

Cellmarker2 raw dataset

Description

A dataset containing marker genes for different cell types from Cellmarker2

Usage

Cellmarker2_raw

Format

A data frame with 20 columns contined in the Cellmarker2 database:

Details

This dataset is used to filter and create a standardized marker list. The dataset can be filtered based on species, tissue class, tissue type, cancer type, and cell type to generate a list of marker genes for specific cell types.

Source

http://117.50.127.228/CellMarker/

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_table, Markers_list_PCTAM, Markers_list_PCTIT, Markers_list_TCellSI, Markers_list_scIBD, PanglaoDB, PanglaoDB_raw, PanglaoDB_table

Cellmarker2 table

Description

A dataset containing marker genes for different cell types from Cellmarker2

Usage

Cellmarker2_table

Format

A list contain different types like species, tissue_class, tissue_type, cancer_type, cell_type

Details

This list is used to choose filters for creation of standardized marker list.

Source

http://117.50.127.228/CellMarker/

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_raw, Markers_list_PCTAM, Markers_list_PCTIT, Markers_list_TCellSI, Markers_list_scIBD, PanglaoDB, PanglaoDB_raw, PanglaoDB_table

Annotate Seurat Object with SlimR Cell Type Predictions

Description

This function assigns SlimR predicted cell types to a Seurat object based on cluster annotations, and stores the results in the meta.data slot.

Usage

Celltype_Annotation(
  seurat_obj,
  cluster_col,
  SlimR_anno_result,
  plot_UMAP = TRUE,
  annotation_col = "Cell_type_SlimR"
)

Arguments

Value

A Seurat object with updated meta.data containing the predicted cell types.

Note

If plot_UMAP = TRUE, this function will print a UMAP plot as a side effect.

See Also

Other Section_3_Automated_Annotation: Celltype_Annotation_PerCell(), Celltype_Calculate(), Celltype_Calculate_PerCell(), Celltype_Verification(), Celltype_Verification_PerCell(), Parameter_Calculate(), percell_workflow

Examples

## Not run: 
sce <- Celltype_Annotation(seurat_obj = sce,
    cluster_col = "seurat_clusters",
    SlimR_anno_result = SlimR_anno_result,
    plot_UMAP = TRUE,
    annotation_col = "Cell_type_SlimR"
    )
    
## End(Not run)

Uses "marker_list" to generate combined plot for cell annotation

Description

Uses "marker_list" to generate combined plot for cell annotation

Usage

Celltype_Annotation_Combined(
  seurat_obj,
  gene_list,
  species,
  cluster_col = "seurat_clusters",
  assay = "RNA",
  save_path = NULL,
  colour_low = "white",
  colour_high = "navy"
)

Arguments

Value

The cell annotation picture is saved in "save_path".

See Also

Other Section_4_Semi_Automated_Annotation: Celltype_Annotation_Features(), Celltype_Annotation_Heatmap()

Examples

## Not run: 
Celltype_Annotation_Combined(seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_Annotation_Combined"),
    colour_low = "white",
    colour_high = "navy"
    )
    
## End(Not run)

Annotate cell types using features plot with different marker databases

Description

This function dynamically selects the appropriate annotation method based on the gene_list_type parameter. It supports marker databases from Cellmarker2, PanglaoDB, Seurat (via FindAllMarkers), or Excel files.

Usage

Celltype_Annotation_Features(
  seurat_obj,
  gene_list,
  gene_list_type = "Default",
  species = NULL,
  cluster_col = "seurat_clusters",
  assay = "RNA",
  save_path = NULL,
  min_counts = 1,
  metric_names = NULL,
  colour_low = "white",
  colour_high = "navy",
  colour_low_mertic = "white",
  colour_high_mertic = "navy",
  ...
)

Arguments

Value

Saves cell type annotation PNGs in save_path. Returns invisibly.

See Also

Other Section_4_Semi_Automated_Annotation: Celltype_Annotation_Combined(), Celltype_Annotation_Heatmap()

Examples

## Not run: 
# Example for Cellmarker2
Celltype_Annotation_Features(seurat_obj = sce,
    gene_list = Markers_list_Cellmarker2,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_annotation_Cellmarker2"),
    colour_low = "white",
    colour_high = "navy",
    colour_low_mertic = "white",
    colour_high_mertic = "navy",
    )

# Example for PanglaoDB
Celltype_Annotation_Features(seurat_obj = sce,
    gene_list = Markers_list_panglaoDB,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_annotation_PanglaoDB")
    colour_low = "white",
    colour_high = "navy",
    colour_low_mertic = "white",
    colour_high_mertic = "navy",
    )

# Example for Seurat marker list
Celltype_Annotation_Features(seurat_obj = sce,
    gene_list = Markers_list_Seurat,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_annotation_Seurat")
    colour_low = "white",
    colour_high = "navy",
    colour_low_mertic = "white",
    colour_high_mertic = "navy",
    )

# Example for Excel marker list
Celltype_Annotation_Features(seurat_obj = sce,
    gene_list = Markers_list_Excel,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_annotation_Excel")
    colour_low = "white",
    colour_high = "navy",
    colour_low_mertic = "white",
    colour_high_mertic = "navy",
    )

## End(Not run)

Uses "marker_list" to generate heatmap for cell annotation

Description

Uses "marker_list" to generate heatmap for cell annotation

Usage

Celltype_Annotation_Heatmap(
  seurat_obj,
  gene_list,
  species,
  cluster_col = "seurat_clusters",
  assay = "RNA",
  min_expression = 0.1,
  specificity_weight = 3,
  colour_low = "navy",
  colour_high = "firebrick3"
)

Arguments

Value

The heatmap of the comparison between "cluster_col" in the Seurat object and the given gene set "gene_list" needs to be annotated.

See Also

Other Section_4_Semi_Automated_Annotation: Celltype_Annotation_Combined(), Celltype_Annotation_Features()

Examples

## Not run: 
Celltype_Annotation_Heatmap(seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    min_expression = 0.1,
    specificity_weight = 3,
    colour_low = "navy",
    colour_high = "firebrick3"
    )
    
## End(Not run)

Annotate Seurat Object with Per-Cell SlimR Predictions

Description

This function assigns SlimR per-cell predicted cell types directly to individual cells in a Seurat object's meta.data slot.

Usage

Celltype_Annotation_PerCell(
  seurat_obj,
  SlimR_percell_result,
  plot_UMAP = TRUE,
  annotation_col = "Cell_type_PerCell_SlimR",
  plot_confidence = FALSE
)

Arguments

Value

A Seurat object with updated meta.data containing:

• annotation_col: Predicted cell type for each cell

• paste0(annotation_col, "_score"): Max score for each cell

• paste0(annotation_col, "_confidence"): Confidence score for each cell

Note

If plot_UMAP = TRUE, this function will print UMAP plot(s) as a side effect.

See Also

Other Section_3_Automated_Annotation: Celltype_Annotation(), Celltype_Calculate(), Celltype_Calculate_PerCell(), Celltype_Verification(), Celltype_Verification_PerCell(), Parameter_Calculate(), percell_workflow

Examples

## Not run: 
# Run per-cell annotation
result <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human"
)

# Annotate Seurat object
sce <- Celltype_Annotation_PerCell(
    seurat_obj = sce,
    SlimR_percell_result = result,
    plot_UMAP = TRUE,
    annotation_col = "Cell_type_PerCell_SlimR"
)

## End(Not run)

Uses "marker_list" to calculate probability, prediction results, AUC and generate heatmap for cell annotation

Description

Uses "marker_list" to calculate probability, prediction results, AUC and generate heatmap for cell annotation

Usage

Celltype_Calculate(
  seurat_obj,
  gene_list,
  species,
  cluster_col = "seurat_clusters",
  assay = "RNA",
  min_expression = 0.1,
  specificity_weight = 3,
  threshold = 0.6,
  compute_AUC = TRUE,
  plot_AUC = TRUE,
  AUC_correction = FALSE,
  colour_low = "navy",
  colour_high = "firebrick3"
)

Arguments

Value

A list containing:

• Expression_list: List of expression matrices for each cell type

• Proportion_list: List of proportion of expression for each cell type

• Expression_scores_matrix: Matrix of expression scores

• Probability_matrix: Matrix of normalized probabilities

• Prediction_results: Data frame with cluster annotations including:

  • cluster_col: Cluster identifier

  • Predicted_cell_type: Primary predicted cell type

  • AUC: Area Under the Curve value (when compute_AUC = TRUE)

  • Alternative_cell_types: Semi-colon separated alternative cell types

• Heatmap_plot: Heatmap visualization of probability matrix (pheatmap object). Can be displayed using print() or plot()

• AUC_plot: AUC visualization of Predicted cell type (ggplot object)

• AUC_list: The resulting list of AUC values calculated for genes in alternative cell types above the approximate threshold

See Also

Other Section_3_Automated_Annotation: Celltype_Annotation(), Celltype_Annotation_PerCell(), Celltype_Calculate_PerCell(), Celltype_Verification(), Celltype_Verification_PerCell(), Parameter_Calculate(), percell_workflow

Examples

## Not run: 
SlimR_anno_result <- Celltype_Calculate(seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    min_expression = 0.1,
    specificity_weight = 3,
    threshold = 0.6,
    compute_AUC = TRUE,
    plot_AUC = TRUE,
    AUC_correction = FALSE,
    colour_low = "navy",
    colour_high = "firebrick3"
    )
    
## End(Not run)

Per-cell annotation using marker expression and optional UMAP spatial smoothing

Description

Unlike cluster-based annotation, this function assigns cell type labels to each individual cell based on marker gene expression profiles. Optionally uses UMAP coordinates to smooth predictions via k-nearest neighbor voting.

Usage

Celltype_Calculate_PerCell(
  seurat_obj,
  gene_list,
  species,
  assay = "RNA",
  method = c("weighted", "mean", "AUCell"),
  min_expression = 0.1,
  use_umap_smoothing = FALSE,
  umap_reduction = "umap",
  k_neighbors = 15,
  smoothing_weight = 0.3,
  min_score = "auto",
  min_confidence = 1.2,
  return_scores = FALSE,
  ncores = 1,
  chunk_size = 5000,
  verbose = TRUE
)

Arguments

Details

Scoring Methods

"weighted" (recommended): Combines normalized expression with detection rate. For each cell and cell type: score = mean(expr_i * weight_i) where weight_i is derived from the marker's specificity across the dataset.

"mean": Simple average of normalized marker expression. Fast but less discriminative for overlapping marker sets.

"AUCell": Rank-based scoring similar to AUCell package. For each cell, genes are ranked by expression, and the score is the proportion of marker genes in the top X% of expressed genes. Robust to technical variation.

UMAP Smoothing

When use_umap_smoothing = TRUE, the function:

1. Computes initial per-cell scores

2. Finds k nearest neighbors in UMAP space for each cell

3. Smooths scores by weighted averaging with neighbors

4. Re-assigns cell types based on smoothed scores

This helps reduce noise and improve consistency of annotations within spatially coherent regions.

Value

A list containing:

• Cell_annotations: Data frame with Cell_barcode, Predicted_cell_type, Max_score, Confidence

• Cell_confidence: Numeric vector of confidence scores per cell

• Summary: Summary table of cell type counts and percentages

• Expression_list: List of mean expression matrices per cell type (for verification)

• Proportion_list: List of detection proportion matrices per cell type

• Prediction_results: Summary data frame with per-cell-type statistics

• Probability_matrix: Full cell × cell_type probability matrix (normalized)

• Raw_score_matrix: Full cell × cell_type raw score matrix (before normalization)

• Parameters: List of parameters used including adaptive thresholds

• Cell_scores: (if return_scores=TRUE) Same as Probability_matrix

See Also

Other Section_3_Automated_Annotation: Celltype_Annotation(), Celltype_Annotation_PerCell(), Celltype_Calculate(), Celltype_Verification(), Celltype_Verification_PerCell(), Parameter_Calculate(), percell_workflow

Examples

## Not run: 
# Basic per-cell annotation
result <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    method = "weighted"
)

# Add annotations to Seurat object
sce$Cell_type_PerCell <- result$Cell_annotations$Predicted_cell_type

# With UMAP smoothing for more consistent annotations
result_smooth <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    use_umap_smoothing = TRUE,
    k_neighbors = 20,
    smoothing_weight = 0.3
)

## End(Not run)

Perform cell type verification and generate the validation dotplot

Description

This function performs verification of predicted cell types by selecting high log2FC and high expression proportion genes and generates and generate the validation dotplot.

Usage

Celltype_Verification(
  seurat_obj,
  SlimR_anno_result,
  assay = "RNA",
  gene_number = 5,
  colour_low = "white",
  colour_high = "navy",
  annotation_col = "Cell_type_SlimR"
)

Arguments

Value

A ggplot object showing expression of top variable genes.

See Also

Other Section_3_Automated_Annotation: Celltype_Annotation(), Celltype_Annotation_PerCell(), Celltype_Calculate(), Celltype_Calculate_PerCell(), Celltype_Verification_PerCell(), Parameter_Calculate(), percell_workflow

Examples

## Not run: 
Celltype_Verification(seurat_obj = sce,
    SlimR_anno_result = SlimR_anno_result,
    assay = "RNA",
    gene_number = 5,
    colour_low = "white",
    colour_high = "navy",
    annotation_col = "Cell_type_SlimR"
    )
    
## End(Not run)

Verify per-cell annotations with marker expression dotplot

Description

This function verifies per-cell SlimR annotations by generating a dotplot showing marker gene expression across predicted cell types.

Usage

Celltype_Verification_PerCell(
  seurat_obj,
  SlimR_percell_result,
  assay = "RNA",
  gene_number = 5,
  colour_low = "white",
  colour_high = "navy",
  annotation_col = "Cell_type_PerCell_SlimR",
  min_cells = 10
)

Arguments

Value

A ggplot object showing marker gene expression dotplot.

See Also

Other Section_3_Automated_Annotation: Celltype_Annotation(), Celltype_Annotation_PerCell(), Celltype_Calculate(), Celltype_Calculate_PerCell(), Celltype_Verification(), Parameter_Calculate(), percell_workflow

Examples

## Not run: 
# After running Celltype_Calculate_PerCell and Celltype_Annotation_PerCell
dotplot <- Celltype_Verification_PerCell(
    seurat_obj = sce,
    SlimR_percell_result = result,
    gene_number = 5,
    annotation_col = "Cell_type_PerCell_SlimR"
)
print(dotplot)

## End(Not run)

Uses "marker_list" from Cellmarker2 for cell annotation

Description

Uses "marker_list" from Cellmarker2 for cell annotation

Usage

Celltype_annotation_Cellmarker2(
  seurat_obj,
  gene_list,
  species,
  cluster_col = "seurat_clusters",
  assay = "RNA",
  save_path = NULL,
  min_counts = 1,
  colour_low = "white",
  colour_high = "navy",
  colour_low_mertic = "white",
  colour_high_mertic = "navy"
)

Arguments

Value

The cell annotation picture is saved in "save_path".

See Also

Other Section_5_Other_Functions_Provided: Celltype_annotation_Excel(), Celltype_annotation_PanglaoDB(), Celltype_annotation_Seurat()

Examples

## Not run: 
Celltype_annotation_Cellmarker2(seurat_obj = sce,
    gene_list = Markers_list_Cellmarker2,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_annotation_Cellmarker2")
    colour_low = "white",
    colour_high = "navy",
    colour_low_mertic = "white",
    colour_high_mertic = "navy",
    )
    
## End(Not run)

Uses "marker_list" from Excel input for cell annotation

Description

Uses "marker_list" from Excel input for cell annotation

Usage

Celltype_annotation_Excel(
  seurat_obj,
  gene_list,
  species,
  cluster_col = "seurat_clusters",
  assay = "RNA",
  save_path = NULL,
  metric_names = NULL,
  colour_low = "white",
  colour_high = "navy",
  colour_low_mertic = "white",
  colour_high_mertic = "navy"
)

Arguments

Value

The cell annotation picture is saved in "save_path".

See Also

Other Section_5_Other_Functions_Provided: Celltype_annotation_Cellmarker2(), Celltype_annotation_PanglaoDB(), Celltype_annotation_Seurat()

Examples

## Not run: 
Celltype_annotation_Excel(seurat_obj = sce,
    gene_list = Markers_list_Excel,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_annotation_Excel")
    colour_low = "white",
    colour_high = "navy",
    colour_low_mertic = "white",
    colour_high_mertic = "navy",
    )
    
## End(Not run)

Uses "marker_list" from PanglaoDB for cell annotation

Description

Uses "marker_list" from PanglaoDB for cell annotation

Usage

Celltype_annotation_PanglaoDB(
  seurat_obj,
  gene_list,
  species,
  cluster_col = "seurat_clusters",
  assay = "RNA",
  save_path = NULL,
  metric_names = NULL,
  colour_low = "white",
  colour_high = "navy",
  colour_low_mertic = "white",
  colour_high_mertic = "navy"
)

Arguments

Value

The cell annotation picture is saved in "save_path".

See Also

Other Section_5_Other_Functions_Provided: Celltype_annotation_Cellmarker2(), Celltype_annotation_Excel(), Celltype_annotation_Seurat()

Examples

## Not run: 
Celltype_annotation_PanglaoDB(seurat_obj = sce,
    gene_list = Markers_list_panglaoDB,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_annotation_PanglaoDB")
    colour_low = "white",
    colour_high = "navy",
    colour_low_mertic = "white",
    colour_high_mertic = "navy",
    )
    
## End(Not run)

Uses "marker_list" from Seurat object for cell annotation

Description

Uses "marker_list" from Seurat object for cell annotation

Usage

Celltype_annotation_Seurat(
  seurat_obj,
  gene_list,
  species,
  cluster_col = "seurat_clusters",
  assay = "RNA",
  save_path = NULL,
  metric_names = NULL,
  colour_low = "white",
  colour_high = "navy",
  colour_low_mertic = "white",
  colour_high_mertic = "navy"
)

Arguments

Value

The cell annotation picture is saved in "save_path".

See Also

Other Section_5_Other_Functions_Provided: Celltype_annotation_Cellmarker2(), Celltype_annotation_Excel(), Celltype_annotation_PanglaoDB()

Examples

## Not run: 
Celltype_annotation_Seurat(seurat_obj = sce,
    gene_list = Markers_list_Seurat,
    species = "Human",
    cluster_col = "seurat_clusters",
    assay = "RNA",
    save_path = file.path(tempdir(),"SlimR_Celltype_annotation_Seurat")
    colour_low = "white",
    colour_high = "navy",
    colour_low_mertic = "white",
    colour_high_mertic = "navy",
    )
    
## End(Not run)

Create Marker_list from the Cellmarkers2 database

Description

Create Marker_list from the Cellmarkers2 database

Usage

Markers_filter_Cellmarker2(
  df,
  species = NULL,
  tissue_class = NULL,
  tissue_type = NULL,
  cancer_type = NULL,
  cell_type = NULL
)

Arguments

Value

The standardized "Marker_list" in the SlimR package

See Also

Other Section_2_Standardized_Markers_List: Markers_filter_PanglaoDB(), Read_excel_markers(), Read_seurat_markers()

Examples

Cellmarker2 <- SlimR::Cellmarker2
Markers_list_Cellmarker2 <- Markers_filter_Cellmarker2(
    Cellmarker2,
    species = "Human",
    tissue_class = "Intestine",
    tissue_type = NULL,
    cancer_type = NULL,
    cell_type = NULL
    )

Create Marker_list from the PanglaoDB database

Description

Create Marker_list from the PanglaoDB database

Usage

Markers_filter_PanglaoDB(df, species_input, organ_input)

Arguments

Value

The standardized "Marker_list" in the SlimR package

See Also

Other Section_2_Standardized_Markers_List: Markers_filter_Cellmarker2(), Read_excel_markers(), Read_seurat_markers()

Examples

PanglaoDB <- SlimR::PanglaoDB
Markers_list_panglaoDB <- Markers_filter_PanglaoDB(
    PanglaoDB,
    species_input = 'Human',
    organ_input = 'GI tract'
    )

List of Macrophage subtype markers in the article "Macrophage diversity in cancer revisited in the era of single-cell omics"

Description

A dataset containing marker genes for different Macrophage subtypes from the article "Macrophage diversity in cancer revisited in the era of single-cell omics"

Usage

Markers_list_PCTAM

Format

A list with 7 tables.

Details

This list is a table of 7 types of Tumor-associated macrophages (TAMs) markers obtained from the article "Macrophage diversity in cancer revisited in the era of single-cell omics". The data source is "https://doi.org/10.1016/j.it.2022.04.008", and the reference literature is: Ruo-Yu Ma et al. (2022) doi:10.1016/j.it.2022.04.008.

Source

doi:10.1016/j.it.2022.04.008

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_raw, Cellmarker2_table, Markers_list_PCTIT, Markers_list_TCellSI, Markers_list_scIBD, PanglaoDB, PanglaoDB_raw, PanglaoDB_table

List of T cell subtype markers in the article "Pan-cancer single cell landscape of tumor-infiltrating T cells"

Description

A dataset containing marker genes for different T cell types from the article "Pan-cancer single cell landscape of tumor-infiltrating T cells"

Usage

Markers_list_PCTIT

Format

A list with 40 tables.

Details

This list is a table of 40 types of pan-cancer tumor-infiltrating T cell (PCTIT) markers obtained from the article "Pan-cancer single cell landscapeof tumor-infiltrating T cells". The data source is "https://doi.org/10.1126/science.abe6474", and the reference literature is: L. Zheng et al. (2021) doi:10.1126/science.abe6474.

Source

doi:10.1126/science.abe6474

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_raw, Cellmarker2_table, Markers_list_PCTAM, Markers_list_TCellSI, Markers_list_scIBD, PanglaoDB, PanglaoDB_raw, PanglaoDB_table

List of T cell subtype markers in the article TCellSI

Description

A dataset containing marker genes for different T cell subtypes from TCellSI

Usage

Markers_list_TCellSI

Format

A list with ten tables.

Details

This list is a table of 10 types of T cell markers obtained from TCellSI. The data source is "https://github.com/GuoBioinfoLab/TCellSI/blob/main/data/markers.rda", and the reference literature is: Yang et al. (2024) doi:10.1002/imt2.231.

Source

https://github.com/GuoBioinfoLab/TCellSI/

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_raw, Cellmarker2_table, Markers_list_PCTAM, Markers_list_PCTIT, Markers_list_scIBD, PanglaoDB, PanglaoDB_raw, PanglaoDB_table

List of cell type markers in the article scIBD

Description

A dataset containing marker genes for different human intestine cell types from scIBD

Usage

Markers_list_scIBD

Format

A list with one hundred and one tables.

Details

This list is a table of 101 types of human intestine cell types markers obtained from scIBD. The article doi source is "https://doi.org/10.1038/s43588-023-00464-9", and the reference literature is: Nie et al. (2023) doi:10.1038/s43588-023-00464-9. Note: The 'Markers_list_scIBD' was generated using section 2.5.2 and the parameters 'sort_by = "logFC"' and 'gene_filter = 20' were set.

Source

doi:10.1038/s43588-023-00464-9

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_raw, Cellmarker2_table, Markers_list_PCTAM, Markers_list_PCTIT, Markers_list_TCellSI, PanglaoDB, PanglaoDB_raw, PanglaoDB_table

PanglaoDB dataset

Description

A dataset containing marker genes for different cell types from PanglaoDB

Usage

PanglaoDB

Format

A data frame with 9 columns:

Details

This dataset is used to filter and create a standardized marker list.'

Source

https://panglaodb.se/

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_raw, Cellmarker2_table, Markers_list_PCTAM, Markers_list_PCTIT, Markers_list_TCellSI, Markers_list_scIBD, PanglaoDB_raw, PanglaoDB_table

PanglaoDB raw dataset

Description

A dataset containing marker genes for different cell types from PanglaoDB

Usage

PanglaoDB_raw

Format

A data frame with 14 columns contined in the PanglaoDB database:

Details

This dataset is used to filter and create a standardized marker list.'

Source

https://panglaodb.se/

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_raw, Cellmarker2_table, Markers_list_PCTAM, Markers_list_PCTIT, Markers_list_TCellSI, Markers_list_scIBD, PanglaoDB, PanglaoDB_table

PanglaoDB table

Description

A dataset containing marker genes for different cell types from PanglaoDB

Usage

PanglaoDB_table

Format

A list contain different types like species, organ, cell type.

Details

This list is used to choose filters for creation of standardized marker list.

Source

https://panglaodb.se/

See Also

Other Section_0_Database: Cellmarker2, Cellmarker2_raw, Cellmarker2_table, Markers_list_PCTAM, Markers_list_PCTIT, Markers_list_TCellSI, Markers_list_scIBD, PanglaoDB, PanglaoDB_raw

Adaptive Parameter Tuning for Single-Cell Data Annotation in SlimR

Description

This function automatically determines optimal min_expression, specificity_weight, and threshold parameters for single-cell data analysis based on dataset characteristics using adaptive algorithms derived from empirical analysis of single-cell datasets.

Usage

Parameter_Calculate(
  seurat_obj,
  features = NULL,
  assay = NULL,
  cluster_col = NULL,
  n_celltypes = 50,
  verbose = TRUE
)

Arguments

Value

A list containing:

• min_expression: Recommended expression threshold

• specificity_weight: Recommended specificity weight

• threshold: Recommended probability threshold for candidate selection

• dataset_features: Extracted dataset characteristics

• parameter_rationale: Explanation of parameter choices

See Also

Other Section_3_Automated_Annotation: Celltype_Annotation(), Celltype_Annotation_PerCell(), Celltype_Calculate(), Celltype_Calculate_PerCell(), Celltype_Verification(), Celltype_Verification_PerCell(), percell_workflow

Examples

## Not run: 
SlimR_params <- Parameter_Calculate(
  seurat_obj = sce,
  features = c("CD3E", "CD4", "CD8A"),
  assay = "RNA",
  cluster_col = "seurat_clusters",
  n_celltypes = 98,
  verbose = TRUE
  )

## End(Not run)

Create "Marker_list" from Excel files ".xlsx"

Description

Create "Marker_list" from Excel files ".xlsx"

Usage

Read_excel_markers(path, has_colnames = TRUE)

Arguments

Value

The standardized "Marker_list" in the SlimR package.

See Also

Other Section_2_Standardized_Markers_List: Markers_filter_Cellmarker2(), Markers_filter_PanglaoDB(), Read_seurat_markers()

Examples

## Not run: 
Markers_list_Excel <- Read_excel_markers(
    "D:/Laboratory/Marker_load.xlsx"
    )

## End(Not run)

Create "Marker_list" from Seurat object

Description

Create "Marker_list" from Seurat object

Usage

Read_seurat_markers(
  df,
  sources = c("Seurat", "presto"),
  sort_by = "FSS",
  gene_filter = 20
)

Arguments

Value

The standardized "Marker_list" in the SlimR package.

See Also

Other Section_2_Standardized_Markers_List: Markers_filter_Cellmarker2(), Markers_filter_PanglaoDB(), Read_excel_markers()

Examples

## Not run: 
# Example for Seurat sources markers
seurat_markers <- Seurat::FindAllMarkers(
    object = sce,
    group.by = "Cell_type",
    only.pos = TRUE)

Markers_list_Seurat <- Read_seurat_markers(seurat_markers,
    sources = "Seurat",
    sort_by = "avg_log2FC",
    gene_filter = 20
    )

# Example for presto sources markers
seurat_markers <- dplyr::filter(
    presto::wilcoxauc(
      X = sce,
      group_by = "Cell_type",
      seurat_assay = "RNA"
      ),
    padj < 0.05, logFC > 0.5
    )

Markers_list_Seurat <- Read_seurat_markers(seurat_markers,
    sources = "presto",
    sort_by = "logFC",
    gene_filter = 20
    )

## End(Not run)

Calculate Cluster Variability (Use in package)

Description

Measures the degree of separation between different cell clusters based on expression patterns.

Usage

calculate_cluster_variability(data.features, features)

Arguments

Value

Numeric value representing cluster separation strength

See Also

Other Section_1_Functions_Use_in_Package: calculate_expression(), calculate_expression_skewness(), calculate_probability(), compute_adaptive_parameters(), estimate_batch_effect(), extract_dataset_features()

Counts average expression of gene set (Use in package)

Description

Counts average expression of gene set (Use in package)

Usage

calculate_expression(
  object,
  features,
  assay = NULL,
  cluster_col = NULL,
  colour_low = "white",
  colour_high = "navy"
)

Arguments

Value

Average expression genes and relatied informations in the input "Seurat" object given "cluster_col" and given "features".

See Also

Other Section_1_Functions_Use_in_Package: calculate_cluster_variability(), calculate_expression_skewness(), calculate_probability(), compute_adaptive_parameters(), estimate_batch_effect(), extract_dataset_features()

Calculate Expression Distribution Skewness (Use in package)

Description

Computes the average skewness of gene expression distributions across all features.

Usage

calculate_expression_skewness(expression_matrix)

Arguments

Value

Mean absolute skewness across all genes

See Also

Other Section_1_Functions_Use_in_Package: calculate_cluster_variability(), calculate_expression(), calculate_probability(), compute_adaptive_parameters(), estimate_batch_effect(), extract_dataset_features()

Calculate gene set expression and infer probabilities with control datasets (Use in package)

Description

Calculate gene set expression and infer probabilities with control datasets (Use in package)

Usage

calculate_probability(
  object,
  features,
  assay = NULL,
  cluster_col = NULL,
  min_expression = 0.1,
  specificity_weight = 3
)

Arguments

Value

Average expression of genes in the input "Seurat" object given "cluster_col" and given "features".

See Also

Other Section_1_Functions_Use_in_Package: calculate_cluster_variability(), calculate_expression(), calculate_expression_skewness(), compute_adaptive_parameters(), estimate_batch_effect(), extract_dataset_features()

Compute Adaptive Parameters Based on Dataset Features (Use in package)

Description

Calculates optimal min_expression, specificity_weight, and threshold parameters using continuous adaptive algorithms based on dataset characteristics.

Usage

compute_adaptive_parameters(dataset_features, n_celltypes = 50)

Arguments

Value

List containing min_expression, specificity_weight, threshold, and rationale

See Also

Other Section_1_Functions_Use_in_Package: calculate_cluster_variability(), calculate_expression(), calculate_expression_skewness(), calculate_probability(), estimate_batch_effect(), extract_dataset_features()

Estimate Batch Effect Strength (Use in package)

Description

Roughly estimates the potential impact of batch effects using available metadata.

Usage

estimate_batch_effect(seurat_obj, assay)

Arguments

Value

Batch effect score (0 indicates no detectable batch effect)

See Also

Other Section_1_Functions_Use_in_Package: calculate_cluster_variability(), calculate_expression(), calculate_expression_skewness(), calculate_probability(), compute_adaptive_parameters(), extract_dataset_features()

Extract Dataset Characteristics for Adaptive Parameter Calculation (Use in package)

Description

Computes various statistical features from single-cell data that are used as input for the parameter prediction model.

Usage

extract_dataset_features(
  seurat_obj,
  features,
  assay = NULL,
  cluster_col = NULL
)

Arguments

Value

List of dataset characteristics including expression statistics, variability measures, and cluster properties

See Also

Other Section_1_Functions_Use_in_Package: calculate_cluster_variability(), calculate_expression(), calculate_expression_skewness(), calculate_probability(), compute_adaptive_parameters(), estimate_batch_effect()

Per-Cell Annotation Workflow Example

Description

Example workflow for using SlimR's per-cell annotation functions

Overview

The per-cell annotation workflow in SlimR provides an alternative to cluster-based annotation by scoring and labeling individual cells based on marker expression. This is useful when:

• Clusters contain mixed cell types

• You want finer-grained annotations

• Cell states exist on a continuum

• UMAP spatial context can improve annotation quality

Basic Workflow

# 1. Prepare your Seurat object (must have normalized data)
library(SlimR)
library(Seurat)

# 2. Create or load marker list
Markers_list <- Markers_filter_Cellmarker2(
    Cellmarker2,
    species = "Human",
    tissue_class = "Intestine"
)

# 3. Run per-cell annotation
result <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    method = "weighted",          # "weighted", "mean", or "AUCell"
    min_expression = 0.1,
    min_score = 0.1,
    verbose = TRUE
)

# 4. Annotate Seurat object
sce <- Celltype_Annotation_PerCell(
    seurat_obj = sce,
    SlimR_percell_result = result,
    plot_UMAP = TRUE,
    plot_confidence = TRUE,
    annotation_col = "Cell_type_PerCell"
)

# 5. Verify annotations
dotplot <- Celltype_Verification_PerCell(
    seurat_obj = sce,
    SlimR_percell_result = result,
    gene_number = 5,
    annotation_col = "Cell_type_PerCell"
)
print(dotplot)

Advanced

UMAP Spatial Smoothing:

# Use UMAP coordinates to smooth predictions via k-NN
# This reduces noise and improves consistency in spatial regions

result_smooth <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    use_umap_smoothing = TRUE,
    k_neighbors = 20,              # Number of neighbors to consider
    smoothing_weight = 0.3,        # 30
    verbose = TRUE
)

# Compare smoothed vs unsmoothed
sce$Cell_type_Smooth <- result_smooth$Cell_annotations$Predicted_cell_type
sce$Cell_type_Raw <- result$Cell_annotations$Predicted_cell_type

DimPlot(sce, group.by = "Cell_type_Raw") | 
  DimPlot(sce, group.by = "Cell_type_Smooth")

Scoring Methods Comparison

# Method 1: Weighted (recommended for most cases)
# Combines expression with marker specificity and detection rate
result_weighted <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    method = "weighted"
)

# Method 2: Mean (simple, fast)
# Just averages normalized marker expression
result_mean <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    method = "mean"
)

# Method 3: AUCell (rank-based, robust to batch effects)
# Scores based on proportion of markers in top 5
result_aucell <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    method = "AUCell"
)

Comparing Cluster vs Per-Cell Annotation

# Cluster-based annotation (original SlimR approach)
cluster_result <- Celltype_Calculate(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    cluster_col = "seurat_clusters"
)

sce <- Celltype_Annotation(
    seurat_obj = sce,
    cluster_col = "seurat_clusters",
    SlimR_anno_result = cluster_result,
    annotation_col = "Cell_type_Cluster"
)

# Per-cell annotation
percell_result <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human"
)

sce <- Celltype_Annotation_PerCell(
    seurat_obj = sce,
    SlimR_percell_result = percell_result,
    annotation_col = "Cell_type_PerCell"
)

# Compare
library(ggplot2)
library(patchwork)

p1 <- DimPlot(sce, group.by = "Cell_type_Cluster") + 
      ggtitle("Cluster-based")
p2 <- DimPlot(sce, group.by = "Cell_type_PerCell") + 
      ggtitle("Per-cell")

p1 | p2

# Check agreement
table(sce$Cell_type_Cluster, sce$Cell_type_PerCell)

Performance Optimization

# For large datasets, adjust chunk_size to manage memory
result <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    chunk_size = 10000,  # Process 10k cells at a time
    verbose = TRUE
)

# For UMAP smoothing, install RANN for 10-100x speedup
# install.packages("RANN")

result_smooth <- Celltype_Calculate_PerCell(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human",
    use_umap_smoothing = TRUE,
    k_neighbors = 15
    # RANN will be used automatically if installed
)

Accessing Results

# Cell-level annotations
head(result$Cell_annotations)
#   Cell_barcode Predicted_cell_type Max_score Confidence
# 1  AAACCTGAG... Enterocyte          0.85      0.62
# 2  AAACCTGCA... Goblet cell         0.72      0.45

# Summary statistics
result$Summary
#   Cell_type       Count Percentage
# 1 Enterocyte      5432  45.2
# 2 Goblet cell     2156  17.9

# Full probability matrix (if return_scores = TRUE)
result$Probability_matrix[1:5, 1:3]
#              Enterocyte Goblet_cell Stem_cell
# AAACCTGAG... 0.85       0.10        0.05

# Extract high-confidence cells
high_conf <- result$Cell_annotations$Cell_barcode[
    result$Cell_annotations$Confidence > 0.5
]

# Extract uncertain cells for manual review
uncertain <- result$Cell_annotations$Cell_barcode[
    result$Cell_annotations$Confidence < 0.2
]

See Also

Other Section_3_Automated_Annotation: Celltype_Annotation(), Celltype_Annotation_PerCell(), Celltype_Calculate(), Celltype_Calculate_PerCell(), Celltype_Verification(), Celltype_Verification_PerCell(), Parameter_Calculate()

Plot Method for pheatmap Objects

Description

This S3 method allows pheatmap objects (returned by Celltype_Calculate()) to be plotted using the generic plot() function. Without this method, attempting to use plot() on a pheatmap object results in an error.

Usage

## S3 method for class 'pheatmap'
plot(x, ...)

Arguments

Details

Pheatmap objects contain a gtable component that needs to be drawn using grid graphics. This method handles that automatically when plot() is called.

Alternative ways to display pheatmaps:

• print(pheatmap_object) - Works natively

• plot(pheatmap_object) - Works after loading SlimR

• grid::grid.draw(pheatmap_object$gtable) - Direct access

Value

Invisibly returns the input pheatmap object after displaying it

Examples

## Not run: 
# After running Celltype_Calculate()
cluster_results <- Celltype_Calculate(
    seurat_obj = sce,
    gene_list = Markers_list,
    species = "Human"
)

# Now both of these work:
print(cluster_results$Heatmap_plot)
plot(cluster_results$Heatmap_plot)

## End(Not run)