NEWS

ComplexMap 2.0.0 (2026-02-19)

[Core Logic] Shifted to a Physical-First landscape philosophy. The network layout now defaults to alpha = 0.75, giving 75% weight to physical protein composition to preserve biological specificity and sub-complex architecture.
[Interactive Layer] Replaced all static plotting functions (visualizeMapDirectLabels, visualizeMapWithLegend) with a high-performance interactive engine based on Cytoscape.js and Shiny.
[Visualization] Added explore() and runComplexMapApp() to launch the interactive viewer.
[S3 Object] Overhauled the print.ComplexMap method to provide a "Systems Biology Dashboard" highlighting Functional Diversity, Annotation Coverage, and Physical Density.
[Clustering] Updated functional clustering in generateNodeAttributes to use Average Linkage and a dynamic scaling factor to force functional diversity and prevent "monochromatic" maps.

[Export] Enhanced toCytoscapeJSON with coordinate scaling and edge pruning to optimize browser performance for large networks.
[Export] exportNetwork now supports GraphML and sanitizes R-specific data types for Cytoscape Desktop compatibility.
[Documentation] Fully documented missing arguments (layout_seed, ...) and updated vignettes to reflect the v2.0 API.

[Metrics] Fixed a critical indexing error in evaluateComplexes where the Maximum Matching Ratio (MMR) would return NA when predicted and reference lists differed in size.
[CRAN] Eliminated non-ASCII characters in complexmap_S3.R and qcComplexList.R to ensure cross-platform portability.
[Architecture] Fixed a corrupt lazy-load database issue by ensuring consistent namespace exports.

[createComplexMap] Added ifRefineCpx parameter (default: FALSE). Step 1 (Complex Refinement) is now conditional, allowing users to process raw input lists without automatic merging.

[generateNodeAttributes] Enforced a strict 1:1 mapping between Primary Function and Node Color. Removed the weighted color blending logic to ensure visual consistency across the functional landscape.
[Logging] Updated verbose messaging to explicitly state refinement status during the workflow.

[Parallelism] Refactored evaluateComplexes to use future.apply instead of parallel::mclapply, enabling reliable multi-core computation on Windows.
[Workflow] Improved summarizeThemes() to accept add_to_object = TRUE, allowing theme metadata to be stored directly within the S3 object.

[Dependencies] Moved clue from Suggests to Imports to ensure the Hungarian algorithm is always available for MMR calculation.
[UI] Fixed a runtime error in the interactive viewer when the color.by argument was null.
[Vignettes] Resolved build failures in vignette_02-workflow.Rmd caused by incorrect internal function calls.

[Benchmarking] Introduced benchmarkParameters() to systematically optimize mergeThreshold values against reference complexes.
[Metrics] Added dice and matching_score similarity options to refineComplexList.
[Quantitative] Added support for a color.by argument in visualization functions to map continuous experimental data to node colors.

[Defaults] Updated the default similarityMethod in refineComplexList to "matching_score".
[Traceability] refineComplexList now returns a mergeMap tibble for reproducibility.
[QC] Overhauled qcComplexList output to provide a formatted diagnostic report instead of raw R output.
[Terminology] Renamed the "overlap" similarity metric to "simpson" for scientific accuracy.

Initial stable release.
Introduced the ComplexMap S3 object for structured proteomics workflows.
Added the high-level createComplexMap() wrapper.
Implemented core analysis suite: summarizeThemes(), queryMap(), and exportNetwork().
Released comprehensive documentation and three introductory vignettes.