Package 'scPAS'

Title: Single-Cell Phenotype-Associated Subpopulation Identifier
Description: Identifies phenotype-associated cell subpopulations from single-cell RNA-seq data by integrating bulk RNA-seq data with phenotype information. This package uses network-regularized sparse regression to quantify the strength of association between each cell and a phenotype (e.g., disease stage, tumor metastasis, treatment response, survival outcomes). Compatible with both Seurat v4 (4.0.0+) and Seurat v5 (5.0.0+). The method supports Gaussian (continuous), binomial (binary), and Cox (survival) regression families. Full cross-platform compatibility (Windows, macOS, Linux).
Authors: Aimin Xie [aut] (Original author), Zaoqu Liu [aut, cre] (ORCID: <https://orcid.org/0000-0002-0452-742X>, Maintainer, bug fixes and optimization)
Maintainer: Zaoqu Liu <[email protected]>
License: GPL-3
Version: 1.0.4
Built: 2026-05-23 08:35:15 UTC
Source: https://github.com/Zaoqu-Liu/scPAS

Help Index

The function of imputaion.

Description

The function of imputaion.

Usage

imputation(obj, assay = "RNA", method = c("KNN", "ALRA"))

Arguments

obj

A seurat object.
assay

The assay for imputation. The default is 'RNA'.
method

The method for imputation. The default is 'RNA'.

Value

A seurat object after imputaion.

A method for imputation of missing values in single cell RNA-sequencing data based on ALRA.

Description

A method for imputation of missing values in single cell RNA-sequencing data based on ALRA.

Usage

imputation_ALRA(obj, assay = "RNA")

Arguments

obj

A seurat object.
assay

The assay for imputation. The default is 'RNA'.

Value

A seurat object after imputaion.

A method for imputation of missing values in single cell RNA-sequencing data based on the average expression value of nearest neighbor cells.

Description

A method for imputation of missing values in single cell RNA-sequencing data based on the average expression value of nearest neighbor cells.

Usage

imputation_KNN(obj, assay = "RNA", LogNormalized = TRUE)

Arguments

obj

A seurat object.
assay

The assay for imputation. The default is 'RNA'.
LogNormalized

Whether the data is LogNormalized.

Value

A seurat object after imputaion.

Preprocess the single-cell raw data using functions in the Seurat package

Description

This function provide a simplified-version of Seurat analysis pipeline for single-cell RNA-seq data. It contains the following steps in the pipeline:

  • Create a Seurat object from raw data.

  • Normalize the count data present in a given assay.

  • Identify the variable features.

  • Scales and centers features in the dataset.

  • Run a PCA dimensionality reduction.

  • Constructs a Shared Nearest Neighbor (SNN) Graph for a given dataset.

  • Identify clusters of cells by a shared nearest neighbor (SNN) modularity optimization based clustering algorithm.

  • Run t-distributed Stochastic Neighbor Embedding (t-SNE) dimensionality reduction on selected features.

  • Runs the Uniform Manifold Approximation and Projection (UMAP) dimensional reduction technique.

Usage

run_Seurat(
  counts,
  project = "Single_Cell",
  min.cells = 400,
  min.features = 200,
  meta.data = NULL,
  normalization.method = "LogNormalize",
  scale.factor = 10000,
  selection.method = "vst",
  resolution = 0.6,
  dims_Neighbors = 1:10,
  dims_TSNE = 1:10,
  dims_UMAP = 1:10,
  verbose = TRUE
)

Arguments

counts

A matrix-like object with unnormalized data with cells as columns and features as rows.
project

Project name for the Seurat object.
min.cells

Include features detected in at least this many cells. Will subset the counts matrix as well. To reintroduce excluded features, create a new object with a lower cutoff.
min.features

Include cells where at least this many features are detected.
meta.data

meta data of single cell data.
normalization.method

Method for normalization.

  • LogNormalize: Feature counts for each cell are divided by the total counts for that cell and multiplied by the scale.factor. This is then natural-log transformed using log1p.

  • CLR: Applies a centered log ratio transformation.

  • RC: Relative counts. Feature counts for each cell are divided by the total counts for that cell and multiplied by the scale.factor. No log-transformation is applied. For counts per million (CPM) set scale.factor = 1e6.

scale.factor

Sets the scale factor for cell-level normalization.
selection.method

How to choose top variable features. Choose one of :

  • vst: First, fits a line to the relationship of log(variance) and log(mean) using local polynomial regression (loess). Then standardizes the feature values using the observed mean and expected variance (given by the fitted line). Feature variance is then calculated on the standardized values after clipping to a maximum (see clip.max parameter).

  • mean.var.plot (mvp): First, uses a function to calculate average expression (mean.function) and dispersion (dispersion.function) for each feature. Next, divides features into num.bin (deafult 20) bins based on their average expression, and calculates z-scores for dispersion within each bin. The purpose of this is to identify variable features while controlling for the strong relationship between variability and average expression.

  • dispersion (disp): selects the genes with the highest dispersion values

resolution

Value of the resolution parameter, use a value above (below) 1.0 if you want to obtain a larger (smaller) number of communities.
dims_Neighbors

Dimensions of reduction to use as input.
dims_TSNE

Which dimensions to use as input features for t-SNE.
dims_UMAP

Which dimensions to use as input features for UMAP.
verbose

Print output.

Value

A Seurat object containing cell-cell similarity network, t-SNE and UMAP representations.

scPAS : A tool for identifying Phenotype-Associated cell Subpopulations from single-cell sequencing data by integrating bulk data

Description

scPAS : A tool for identifying Phenotype-Associated cell Subpopulations from single-cell sequencing data by integrating bulk data

Usage

scPAS(
  bulk_dataset,
  sc_dataset,
  phenotype,
  assay = "RNA",
  tag = NULL,
  nfeature = NULL,
  do_imputation = TRUE,
  imputation_method = c("KNN", "ALRA"),
  alpha = NULL,
  network_class = c("SC", "bulk"),
  independent = TRUE,
  family = c("gaussian", "binomial", "cox"),
  permutation_times = 2000,
  FDR.threshold = 0.05,
  n_cores = 1
)

Arguments

bulk_dataset

Matrix. Bulk expression matrix of related disease. Each row represents a gene and each column represents a sample. The input expression values are continuous, such as microarray fluorescent units in logarithmic scale, RNA-seq log-CPMs, log-RPKMs or log-TPMs.
sc_dataset

Matrix or seurat object. Single-cell RNA-seq expression matrix of related disease. Each row represents a gene and each column represents a sample. A Seurat object that contains the preprocessed data and constructed network is preferred. Otherwise, a cell-cell similarity network is constructed based on the input matrix.Otherwise, the raw count expression matrix will be processed by using Seurat's default parameters. See run_Seurat for details.
phenotype

Phenotype annotation of each bulk sample. It can be a continuous dependent variable, binary group indicator vector, or clinical survival data:

  • Continuous dependent variable. Should be a quantitative vector for family = gaussian.

  • Binary group indicator vector. Should be either a 0-1 encoded vector or a factor with two levels for family = binomial.

  • Clinical survival data. Should be a two-column matrix with columns named 'time' and 'status'. The latter is a binary variable, with '1' indicating event (e.g.recurrence of cancer or death), and '0' indicating right censored. The function Surv() in package survival produces such a matrix.

assay

Name of Assay to get.
tag

Names for each phenotypic group. Used for logistic regressions only.
nfeature

Numeric. The Number of features to select as top variable features in sc_dataset. Top variable features will be used to intersect with the features of bulk_dataset. Default is NULL.All features will be used.
do_imputation

Logical. Whether to perform imputation on single-cell data (default: TRUE).
imputation_method

Character. Name of alternative method for imputation.
alpha

Numeric. Parameter used to balance the effect of the l1 norm and the network-based penalties. It can be a number or a searching vector. If alpha = NULL, a default searching vector is used. The range of alpha is in [0,1]. A larger alpha lays more emphasis on the l1 norm.
network_class

The source of feature-feature similarity network. By default this is set to sc and the other one is bulk.
independent

Logical. The background distribution of risk scores is constructed independently of each cell.
family

Character. Response type for the regression model. It depends on the type of the given phenotype and can be family = gaussian for linear regression, family = binomial for classification, or family = cox for Cox regression.
permutation_times

Integer. Number of permutation iterations for statistical significance testing (default: 2000). Higher values increase accuracy but also computation time. Recommended: 1000-5000. For faster testing, use 500-1000.
FDR.threshold

Numeric. FDR value threshold for identifying phenotype-associated cells. The default is 0.05.
n_cores

Integer. Number of CPU cores to use for parallel permutation test (default: 1 for sequential processing). Setting n_cores > 1 enables parallel computing which can significantly speed up the analysis (2-4x faster with 4 cores). Requires 'future' and 'future.apply' packages.

Value

This function returns a Seurat object with the following components added to :

scPAS_para

A list contains the final model parameters added to misc.
PAS result

A data frame containing risk scores (scPAS_RS), normalized risk scores (scPAS_NRS), p-value (scPAS_Pvalue) , adjusted p-value (scPAS_FDR) cell classification labels (scPAS) added to metaData.

scPAS.prediction: A function that uses the scPAS model to make predictions on independent data

Description

scPAS.prediction: A function that uses the scPAS model to make predictions on independent data

Usage

scPAS.prediction(
  model,
  test.data,
  assay = "RNA",
  FDR.threshold = 0.05,
  do_imputation = FALSE,
  imputation_method = "KNN",
  independent = TRUE,
  permutation_times = 2000,
  n_cores = 1
)

Arguments

model

Seurat object. A Seurat object containing the scPAS model (from running scPAS()).
test.data

Matrix or Seurat object. Single-cell RNA-seq expression matrix of related disease. Each row represents a gene and each column represents a sample. A Seurat object that contains the preprocessed data and constructed network is preferred.
assay

Name of Assay to get.
FDR.threshold

Numeric. FDR value threshold for identifying phenotype-associated cells. The default is 0.05.
do_imputation

Logical. Whether to perform imputation on the test data (default: FALSE).
imputation_method

Character. Imputation method: "KNN" or "ALRA".
independent

Logical. Whether to compute background distribution independently for each cell.
permutation_times

Integer. Number of permutations for significance testing (default: 2000).
n_cores

Integer. Number of CPU cores for parallel processing (default: 1).

Value

A seurat object or data frame containing the forecast results.

A function compute the correlation of a sparse matrix.

Description

A function compute the correlation of a sparse matrix.

Usage

sparse.cor(x)

Arguments

x

Matrix. Normalized single cell expression profile extracted from Seurat object.

Value

A correlation matrix.