Skip to contents

Changelog

Source: NEWS.md

CellOracleR 0.1.0

Initial release of CellOracleR, a complete R implementation of the CellOracle Python package for in silico gene perturbation analysis.

Core Features

Gene Regulatory Network Inference

  • Ridge regression with bootstrap aggregation (bagging)
  • Cluster-specific or whole-data GRN fitting
  • Configurable regularization strength

Perturbation Simulation

  • Signal propagation through GRN coefficient matrix
  • Support for knockouts, knockdowns, and overexpression
  • Iterative propagation with configurable depth

Cell State Transition Analysis

  • Transition probability estimation from expression correlation
  • Embedding-based velocity calculation
  • Grid-based flow visualization

Motif Analysis

  • Peak-to-gene association
  • TF binding site scanning via TFBSTools/motifmatchr
  • JASPAR motif database integration

Network Analysis

  • Network centrality metrics (degree, betweenness, eigenvector)
  • Hub gene identification
  • Cluster-specific network comparison

Visualization

  • ggplot2-based plotting functions
  • Vector field visualization
  • Expression change heatmaps

Technical Features

  • Seurat V4/V5 compatibility: Automatic detection and handling
  • C++ acceleration: Core computations via Rcpp/RcppArmadillo
  • Parallel computing: future framework for cross-platform parallelization
  • Cross-platform: Tested on macOS, Linux, and Windows

Dependencies

Required

  • R (>= 4.0.0)
  • Seurat (>= 4.0.0)
  • R6, Rcpp, RcppArmadillo
  • glmnet, igraph, ggplot2

Optional (for motif analysis)

  • TFBSTools, motifmatchr
  • BSgenome, JASPAR2020

References

Kamimoto K, et al. (2023). Dissecting cell identity via network inference and in silico gene perturbation. Nature, 614(7949), 742-751.