SecAct: Secreted Protein Activity Inference

SecAct is an R package designed for inferring the intercellular signaling activity of secreted proteins from gene expression profiles. Users can input multiple modalities of expression data, including spatial, single-cell, or bulk transcriptomics data. The outputs are the inferred signaling activities of 1,170 secreted proteins for each spatial spot, individual cell, or sample, depending on the input data type. Based on the inferred activities, SecAct provides multiple downstream application modules. For spatial data, SecAct can infer the signaling pattern and signaling velocity for secreted proteins. For single-cell data, SecAct can infer the intercellular communication network and signaling flow from source cells to receiver cells. For bulk data, SecAct can infer secreted protein risk scores for a large cohort linked to clinical data, and can infer secreted protein activities that are differentially regulated between two phenotypes. Check the tutorials below for details on how to use the package.

Installation

To install SecAct R package, we recommend using devtools:

# install.packages("devtools")
devtools::install_github("data2intelligence/SecAct")

Or user can install SecAct R package from the source code. Click here to download it.

# install SecAct dependencies
remotes::install_deps("Path_to_the_source_code", force = TRUE)

# install SecAct R package
install.packages("Path_to_the_source_code", repos = NULL, type="source")

🔷 Optional accelerators

After installing the SecAct R package, you can complete all analyses in the tutorial sections. Note that the activity inference functions (SecAct.activity.inference and its *.ST and *.scRNAseq variants) rely on ridge regression and permutation in R language, and may run relatively slowly. If you would like to speed up these functions, please visit here to install the accelerator R package additionally.

🔶 Alternative installation options

1. If you are not familiar with R, you can use the Python version of SecAct available here.

2. If you prefer not to install SecAct locally, a Docker image is available for both the R and Python versions:

# Pull the Docker image.
docker pull psychemistz/secactpy:with-r

3. An online server is also available here. Please note that it currently supports the basic activity inference function only and is limited to datasets with fewer than 10 samples.

Dependencies

• R version >= 4.2.0.
• R packages: Matrix, ggplot2, reshape2, patchwork, NMF, akima, gganimate, metap, circlize, ComplexHeatmap, ggalluvial, networkD3, survival, survminer.

Example

library(SecAct)

dataPath <- file.path(system.file(package = "SecAct"), "extdata/")
expr.diff <- read.table(paste0(dataPath, "Ly86-Fc_vs_Vehicle_logFC.txt"))

# infer activity; ~2 mins
res <- SecAct.activity.inference(inputProfile=expr.diff, is.differential=TRUE) 

head(res$zscore)

Tutorial

SecAct is applicable to multiple modalities of gene expression profiles, including spatial, single-cell, and bulk transcriptomics data.

Spatial transcriptomcis (ST) data

• Signaling patterns and velocities for multi-cellular ST data
  
• Intercellular communication for single-cell resolution ST data

Single-cell RNA sequencing data

• Secreted protein signaling activity for distinct cell states
• Cell-cell communication mediated by secreted proteins

Bulk RNA sequencing data

• Secreted protein signaling activity change between two phenotypes
• Clinical relevance of secreted proteins in a large patient cohort

Miscellaneous

• Activity inference accelerators and batch mode

Contact

For questions, bug reports, or feature requests, please submit an issue. To keep the issue tracker focused and constructive, advertising or promotional content is not permitted.

Citation

Beibei Ru, Lanqi Gong, Emily Yang, Seongyong Park, George Zaki, Kenneth Aldape, Lalage Wakefield, Peng Jiang. Inference of secreted protein activities in intercellular communication. Nature Methods, 2026. [Full Text]