### Install Python Dependencies

Source: https://github.com/gunsagargulati/cytotrace/blob/master/README.md

Install required Python packages for multi-batch analysis.

```shell
$ pip install scanoramaCT update
$ pip install numpy
```

```shell
$ git clone https://github.com/gunsagargulati/scanoramaCT.git
$ cd scanorama/
$ python setup.py install --user
```

--------------------------------

### Install and Load CytoTRACE

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Install the package from GitHub using devtools and load it into the R environment.

```r
# Install from GitHub
install.packages("devtools")
devtools::install_github("gunsagargulati/CytoTRACE")

# Load the package
library(CytoTRACE)
```

--------------------------------

### Install CytoTRACE R Package

Source: https://github.com/gunsagargulati/cytotrace/blob/master/README.md

Install the development version of the package from GitHub using devtools.

```r
install.packages("devtools")
devtools::install_github("gunsagargulati/CytoTRACE")
```

--------------------------------

### Access Included Datasets

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Load example bone marrow differentiation datasets provided with the package for testing.

```r
library(CytoTRACE)

# 10x Genomics bone marrow expression data
data(marrow_10x_expr)
# Format: Matrix with genes as rows, cells as columns
# Values: Unique molecular identifiers (UMIs)
dim(marrow_10x_expr)  # genes x cells

# 10x Genomics bone marrow phenotype labels
data(marrow_10x_pheno)
# Format: Named character vector matching cell IDs
table(marrow_10x_pheno)  # View cell type distribution

# Smart-seq2 bone marrow expression data
data(marrow_plate_expr)
# Format: Matrix with genes as rows, cells as columns
# Values: Raw counts

# Smart-seq2 bone marrow phenotype labels
data(marrow_plate_pheno)
# Format: Named character vector matching cell IDs

# Complete workflow example
results <- CytoTRACE(marrow_10x_expr)
plotCytoTRACE(results, phenotype = marrow_10x_pheno, gene = "Kit")
plotCytoGenes(results, numOfGenes = 15)
```

--------------------------------

### Configure Python Path for Reticulate

Source: https://github.com/gunsagargulati/cytotrace/blob/master/README.md

Set the environment variable to ensure R uses the correct Python installation.

```python
Sys.setenv(RETICULATE_PYTHON="/PATH/TO/PYTHON/BIN"). 
```

--------------------------------

### Initialize Visualization

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Load the library to prepare for generating 2D visualization plots and tables.

```r
library(CytoTRACE)
```

--------------------------------

### Run iCytoTRACE Integration

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Integrate multiple heterogeneous scRNA-seq datasets using Scanorama before computing differentiation predictions.

```r
library(CytoTRACE)

# Load two bone marrow datasets profiled on different platforms
data(marrow_10x_expr)     # 10x Genomics platform
data(marrow_plate_expr)   # Smart-seq2 platform

# Create a list of gene expression matrices
# Each matrix should have genes as rows and cells as columns
datasets <- list(marrow_10x_expr, marrow_plate_expr)

# Run integrated CytoTRACE across multiple batches
results <- iCytoTRACE(datasets)

# Access integrated results
# CytoTRACE scores across all datasets (harmonized)
integrated_cytotrace <- results$CytoTRACE

# Batch-corrected gene expression matrix (Scanorama-corrected)
corrected_expr <- results$exprMatrix

# Integrated gene counts signature
integrated_gcs <- results$GCS

# Corrected gene counts per cell
corrected_counts <- results$Counts

# Genes associated with iCytoTRACE
cyto_genes <- results$cytoGenes

# Low-dimensional embedding coordinates (100 t-SNE components)
embedding_coords <- results$coord

# Filtered cells from all datasets
filtered <- results$filteredCells

# Run with fast mode for large integrated datasets (>10,000 cells)
results_fast <- iCytoTRACE(
  datasets,
  enableFast = TRUE,
  ncores = 4,
  subsamplesize = 1000
)
```

--------------------------------

### Run iCytoTRACE on Multiple Batches

Source: https://github.com/gunsagargulati/cytotrace/blob/master/README.md

Integrate and analyze multiple scRNA-seq datasets from different platforms.

```r
datasets <- list(marrow_10x_expr, marrow_plate_expr)
results <- iCytoTRACE(datasets)
```

--------------------------------

### Run CytoTRACE Analysis and Visualization

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Execute the CytoTRACE algorithm and generate various visualizations including phenotype labels, custom embeddings, and gene expression plots.

```r
data(marrow_10x_expr)
data(marrow_10x_pheno)  # Phenotype labels for cells
results <- CytoTRACE(marrow_10x_expr)

# Basic visualization - just CytoTRACE scores
plotCytoTRACE(results)
# Output files: CytoTRACE_plot.pdf, CytoTRACE_plot_table.txt

# Full visualization with phenotype labels and gene expression
plotCytoTRACE(
  cyto_obj = results,
  phenotype = marrow_10x_pheno,  # Named character vector of cell phenotypes
  gene = "Kit"                    # Gene to visualize (case-insensitive)
)
# Output files:
#   - CytoTRACE_plot.pdf: t-SNE plots colored by CytoTRACE, phenotype, and gene expression
#   - CytoTRACE_by_Phenotype_Boxplot.pdf: Boxplot of CytoTRACE by phenotype
#   - CytoTRACE_plot_table.txt: Tab-delimited table with all values

# Use custom 2D embedding (e.g., UMAP from Seurat)
custom_embedding <- data.frame(
  UMAP1 = runif(ncol(results$exprMatrix)),  # Replace with actual UMAP coordinates
  UMAP2 = runif(ncol(results$exprMatrix))
)
rownames(custom_embedding) <- colnames(results$exprMatrix)

plotCytoTRACE(
  cyto_obj = results,
  phenotype = marrow_10x_pheno,
  gene = "Cd34",
  emb = custom_embedding  # Custom embedding coordinates
)

# Custom color palette
plotCytoTRACE(
  cyto_obj = results,
  phenotype = marrow_10x_pheno,
  colors = c("purple", "yellow", "green")  # Custom gradient colors
)

# Specify output directory
plotCytoTRACE(
  cyto_obj = results,
  phenotype = marrow_10x_pheno,
  gene = "Kit",
  outputDir = "/path/to/output/"  # Must end with "/"
)

# Plot custom values instead of gene expression
custom_values <- results$GCS  # Gene counts signature
plotCytoTRACE(
  cyto_obj = results,
  otherValue = custom_values,
  otherName = "GCS Score"
)
```

--------------------------------

### Run CytoTRACE on Custom Dataset

Source: https://github.com/gunsagargulati/cytotrace/blob/master/README.md

Execute CytoTRACE on a single-cell expression matrix, with optional multi-core support.

```r
results <- cytoTRACE(marrow_10x_expr)
```

```r
results <- cytoTRACE(marrow_10x_expr, enableFast = TRUE, ncores = 8)
```

--------------------------------

### iCytoTRACE Function

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Integrates and analyzes multiple heterogeneous scRNA-seq batches/datasets from different platforms.

```APIDOC
## iCytoTRACE(datasets, enableFast = FALSE, ncores = NULL, subsamplesize = 1000)

### Description
Uses Scanorama to merge datasets before computing integrated CytoTRACE predictions. Requires Python dependencies (scanoramaCT and numpy).

### Parameters
#### Request Body
- **datasets** (list) - Required - List of gene expression matrices (genes as rows, cells as columns).
- **enableFast** (boolean) - Optional - Enable subsampling for large datasets (>10,000 cells).
- **ncores** (integer) - Optional - Number of CPU cores for parallel processing.
- **subsamplesize** (integer) - Optional - Subsample size per chunk for fast mode.

### Response
- **CytoTRACE** (numeric vector) - Harmonized differentiation scores.
- **exprMatrix** (matrix) - Batch-corrected expression matrix.
- **GCS** (numeric vector) - Integrated gene counts signature.
- **Counts** (numeric vector) - Corrected gene counts per cell.
- **cytoGenes** (vector) - Genes associated with iCytoTRACE.
- **coord** (matrix) - Low-dimensional embedding coordinates (100 t-SNE components).
- **filteredCells** (vector) - Filtered cells from all datasets.
```

--------------------------------

### Visualize CytoTRACE Gene Correlations

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Generate bar plots showing genes associated with stemness and differentiation. Requires a previously computed CytoTRACE object.

```r
library(CytoTRACE)

# Run CytoTRACE first
data(marrow_10x_expr)
results <- CytoTRACE(marrow_10x_expr)

# Basic usage - plot top 10 genes on each end
plotCytoGenes(results)
# Output: CytoGenes.pdf - bar plot showing top stemness and differentiation genes

# Show more genes
plotCytoGenes(
  cyto_obj = results,
  numOfGenes = 20  # Show top 20 genes on each end
)

# Custom colors
plotCytoGenes(
  cyto_obj = results,
  numOfGenes = 15,
  colors = c("red", "blue")  # First: less differentiated, Second: more differentiated
)

# Specify output directory
plotCytoGenes(
  cyto_obj = results,
  numOfGenes = 10,
  outputDir = "/path/to/output/"  # Must end with "/"
)

# Manually inspect gene correlations
# Genes positively correlated = associated with less differentiated cells
# Genes negatively correlated = associated with more differentiated cells
cyto_genes <- results$cytoGenes
top_stemness <- head(cyto_genes, 10)      # Most correlated with stemness
top_differentiated <- tail(cyto_genes, 10) # Most correlated with differentiation
```

--------------------------------

### Run CytoTRACE Analysis

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Process gene expression matrices to predict differentiation states, including options for batch correction and fast processing for large datasets.

```r
library(CytoTRACE)

# Load example bone marrow 10x scRNA-seq data (included with package)
# mat: gene expression matrix with genes as rows, cells as columns
data(marrow_10x_expr)

# Basic usage - run CytoTRACE on a single dataset
results <- CytoTRACE(marrow_10x_expr)

# Access results
# CytoTRACE scores: 1.0 = least differentiated (stem-like), 0.0 = most differentiated
cytotrace_scores <- results$CytoTRACE
head(cytotrace_scores)
# Output: Named numeric vector with values between 0 and 1

# Gene counts per cell
gene_counts <- results$Counts

# Gene counts signature (GCS) - geometric mean of top 200 genes correlated with gene counts
gcs_scores <- results$GCS

# Genes correlated with CytoTRACE (sorted by correlation)
cyto_genes <- results$cytoGenes
top_stemness_genes <- head(cyto_genes, 20)  # Genes associated with less differentiated cells

# Normalized expression matrix after processing
expr_matrix <- results$exprMatrix

# Check filtered cells (poor quality cells removed)
filtered_cells <- results$filteredCells

# Run with batch correction (if cells come from different technical batches)
batch_labels <- c(rep("batch1", 500), rep("batch2", 500))  # Example batch vector
results_batch_corrected <- CytoTRACE(marrow_10x_expr, batch = batch_labels)

# Fast mode for large datasets (>3,000 cells) with parallel processing
results_fast <- CytoTRACE(
  marrow_10x_expr,
  enableFast = TRUE,      # Enable subsampling for speed
  ncores = 8,             # Use 8 CPU cores
  subsamplesize = 1000    # Subsample size per chunk
)
```

--------------------------------

### CytoTRACE Function

Source: https://context7.com/gunsagargulati/cytotrace/llms.txt

Predicts differentiation states of single cells from a single scRNA-seq dataset.

```APIDOC
## CytoTRACE(mat, batch = NULL, enableFast = FALSE, ncores = NULL, subsamplesize = 1000)

### Description
Processes gene expression matrices, normalizes data, calculates gene counts signatures, and applies diffusion-based smoothing to generate robust differentiation predictions.

### Parameters
#### Request Body
- **mat** (matrix) - Required - Gene expression matrix with genes as rows and cells as columns.
- **batch** (vector) - Optional - Batch labels for batch correction.
- **enableFast** (boolean) - Optional - Enable subsampling for large datasets (>3,000 cells).
- **ncores** (integer) - Optional - Number of CPU cores for parallel processing.
- **subsamplesize** (integer) - Optional - Subsample size per chunk for fast mode.

### Response
- **CytoTRACE** (numeric vector) - Differentiation scores (1.0 = stem-like, 0.0 = differentiated).
- **Counts** (numeric vector) - Gene counts per cell.
- **GCS** (numeric vector) - Gene counts signature.
- **cytoGenes** (vector) - Genes correlated with CytoTRACE.
- **exprMatrix** (matrix) - Normalized expression matrix.
- **filteredCells** (vector) - List of filtered cells.
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.