Changelog

Source: NEWS.md

CoPro 1.0.0

Major changes

• First public release accompanying the CoPro manuscript.
• Added computeRegressionGeneScores() for regression-based gene weights, which avoids collinearity issues present in PCA back-projection.
• Added copro_download_data() for easy download of example datasets via piggyback.
• Corrected PCA back-projection formula in computeGeneAndCellScores(): gene weights now use 1/sdev instead of sdev.
• Changed transfer_scores() default gs_weight_threshold from 0.005 to 0.
• Added support for regression-based score transfer via getTransferCellScores(..., gene_score_type = "regression").

New vignettes

• “Within-cell-type spatial patterns (Organoid)” – single cell type
• “Cross-cell-type co-progression (Brain MERFISH)” – two cell types
• “Cross-cell-type co-progression with orthogonal axes (Colon Day 3)”
• “Multi-slide analysis and score transfer (Colon Day 9)”
• “Supervised detection of spatial gradients (Kidney)”

Documentation

• Full pkgdown website with grouped function reference.
• Cleaned reproducibility scripts in scripts/ directory.
• Example datasets available via copro_download_data().

CoPro 0.6.1

• Internal refactoring of cell scores, distance matrices, and kernel matrices for cleaner API.
• Added getCellScoresInSitu(), getDistMat(), getKernelMatrix() accessor functions.
• Added self-correlation methods: computeSelfDistance(), computeSelfKernel(), computeSelfBidirCorr().
• Added computeBidirCorrelation() for bidirectional correlation.
• Added testGeneGLM() for gene-level GLM testing.
• Added resample_spatial() and generate_toroidal_permutations().