Skip to contents

scPAS: Single-Cell Phenotype-Associated Subpopulation Identifier

Overview

scPAS is a computational framework for identifying phenotype-associated cell subpopulations from single-cell RNA sequencing (scRNA-seq) data through the integration of bulk transcriptomic profiles and clinical phenotypes. The methodology employs network-regularized sparse regression to quantify the strength of association between individual cells and phenotypic outcomes, enabling both quantitative scoring and statistical inference at single-cell resolution.

scPAS Workflow

Methodological Framework

Statistical Model

scPAS implements the Augmented and Penalized Minimization with L0 (APML0) algorithm, which combines:

  • L1 regularization (LASSO) for sparsity
  • Laplacian regularization for incorporating gene network structure
  • Cross-validation for optimal parameter selection

Supported Regression Families

Family Phenotype Type Application
Gaussian Continuous Age, BMI, gene expression levels
Binomial Binary Case-control, treatment response
Cox Time-to-event Overall survival, progression-free survival

Statistical Inference

Significance is assessed through permutation testing with FDR correction (Benjamini-Hochberg). Cells are classified as:

  • scPAS+: Positively associated (risk score > 0, FDR < threshold)
  • scPAS-: Negatively associated (risk score < 0, FDR < threshold)
  • Non-significant: FDR >= threshold

Installation 

install.packages("scPAS", repos = c(
  "https://zaoqu-liu.r-universe.dev",
  "https://cloud.r-project.org"
))

From GitHub 

if (!require("devtools")) install.packages("devtools")
devtools::install_github("Zaoqu-Liu/scPAS")

Dependencies 

# Bioconductor dependency
if (!require("BiocManager")) install.packages("BiocManager")
BiocManager::install("preprocessCore")

# Optional: parallel computing
install.packages(c("future", "future.apply"))

Documentation

Comprehensive documentation: https://zaoqu-liu.github.io/scPAS/

Vignette Description
Quick Start Basic usage and workflow
Algorithm Mathematical methodology
Visualization Publication-quality figures
Survival Analysis Cox regression application
Binary Classification Treatment response prediction

Quick Start 

library(scPAS)
library(Seurat)

# Run scPAS analysis
result <- scPAS(
  bulk_dataset = bulk_expression,
  sc_dataset = seurat_object,
  phenotype = phenotype_vector,
  family = "gaussian",
  nfeature = 3000,
  permutation_times = 1000,
  n_cores = 4
)

# Extract significant cells
significant_cells <- subset(result, subset = scPAS_FDR < 0.05)
table(significant_cells$scPAS)

Output Structure

Column Description
scPAS_RS Raw risk score
scPAS_NRS Normalized risk score (Z-statistic)
scPAS_Pvalue Permutation-based p-value
scPAS_FDR Benjamini-Hochberg adjusted p-value
scPAS Cell classification (scPAS+/scPAS-/0)

Key Features

  • Multi-modal Integration: Bridges bulk and single-cell transcriptomics
  • Network Regularization: Incorporates gene-gene co-expression structure
  • Flexible Phenotypes: Supports continuous, binary, and survival outcomes
  • Scalable Computation: Parallel processing for large-scale datasets
  • Seurat Integration: Native support for Seurat v4 objects

Citation

If you use scPAS in your research, please cite:

Xie A, Wang H, Zhao J, Wang Z, Xu J, Xu Y. scPAS: single-cell phenotype-associated subpopulation identifier. Briefings in Bioinformatics. 2024;26(1):bbae655. DOI: 10.1093/bib/bbae655

Authors

License

GPL-3.0 | GNU General Public License v3.0