Skip to contents

QUESTIONS

QUESTIONS (Quantifying Expression Spillover Trajectories in Omic Network Series) is an R package for modeling how upstream gene signals propagate through time-ordered biological networks — especially those derived from WGCNA or other coexpression analyses.

QUESTIONS extends the spillover propagation framework to help researchers quantify how biological influence shifts over time, identify key signal-carrying modules, and interpret dynamic omics data with network-level context.

Features

  • Spillover modeling between initiator genes and their neighbors
  • Timecourse analysis across a list of network snapshots
  • Network pruning via thresholds or top-k neighbors
  • Spillover thresholding informed by network structure
  • Sample-level scoring based on spillover-weighted expression
  • Module-level summaries across timepoints
  • Native support for sparse matrices and WGCNA objects

Installation

Install the development version of QUESTIONS from GitHub with pak:

# install.packages("pak")
pak::pak("CarlyBobak/QUESTIONS")

Basic Example

To compute spillover:

  1. Create a similarity matrix (e.g., TOM or adjacency matrix).
  2. Choose one or more initiator genes to seed the propagation.
  3. Compute spillover using calculate_spillover() with optional pruning or normalization.
  4. Interpret or summarize results at the gene, sample, or module level.

For example, you might start with a 2×2 matrix of gene similarity values, initialize spillover from one gene, and observe how signal distributes across the network. Larger networks and timecourse inputs can be handled using spillover_timecourse().

Sample- and module-level summaries can be generated using summarize_spillover_per_sample() or summarize_spillover_by_sample_and_module().

library(QUESTIONS)
library(Matrix)

mat <- Matrix(c(1, 0.3, 0.3, 1), nrow = 2, sparse = TRUE)
rownames(mat) <- colnames(mat) <- c("Gene1", "Gene2")

calculate_spillover(
  network = mat,
  initiator_genes = "Gene1",
  alpha = 0.9
)

Citation

If you use this package, please cite:

Carly Bobak, Jaini Shah, James O’Malley (2025).
QUESTIONS: Quantifying Expression Spillover Trajectories in Omic Network Series.
R package version 0.1.0.

Author & Development

Developed by Carly Bobak in Research Computing and Data, Jaini Shah, and James O’Malley in The Dartmouth Institute at Dartmouth College to support time-resolved and network-informed omics analysis.

Questions and contributions welcome via GitHub Issues.

License

MIT License © 2025 Carly Bobak