NEWS

---

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.