Changelog

Source: NEWS.md

HDAnalyzeR 1.1.0

General Updates

• Improved Function Naming:
  All functions now start with hd_ for consistency and easier searching in RStudio, especially when using multiple packages simultaneously.

• Heatmap Updates:
  Heatmaps are now converted to ggplot objects using ggplotify, enhancing flexibility and uniformity. Now all plots are ggplot objects.

• Enhanced Vignettes:
  Vignettes have been streamlined by suppressing unnecessary outputs and messages.

• Bug Fixes & Warnings:
  Addressed several bugs and added informative warnings and error handling to guide users when potential issues arise.

• Preprocess Module Removal:
  The preprocess module was removed due to redundancy and limited functionality.

Utility Functions

• New Features:
  • hd_initialize(): Initializes a base object used across all functions for simplified and consistent workflows.
    
  • hd_detect_vartype(): Identifies if a variable is categorical or continuous.
  • hd_bin_columns(): Bins continuous variables into categories.
  • hd_filter_by_sex(): Filters data based on sex and updates the data and metadata automatically
  • hd_log_transform(): Log-transforms data.
  • hd_long_data(): Converts wide data to long format.
• Simplified Usage:
  • hd_save_data() (formerly save_df()) now has more user-friendly arguments, aligning with hd_import_data() (formerly import_df()).
  • hd_widen_data() supports custom column selection for wide data creation with arguments like exclude, names_from, and values_from.
• Deprecated Functions:
  generate_df() was deprecated as it is now redundant.

Dimensionality Reduction

• Bug Fixes:
  Fixed an issue in hd_auto_pca() (formerly do_pca()) when handling more than 9 principal components.

Palettes & Themes

• New Functions:
  • hd_show_palettes(): Displays all available palettes in the package.
  • Added selected palettes from the ggsci package.

Imputation

• New Features:
  • hd_na_search(): Summarizes NA distributions in heatmaps and supports user-defined metadata annotations.
    
  • hd_omit_na(): Removes rows with NA values based on specified columns.
• Removed Features:
  impute_mice() was removed for simplicity. Users requiring advanced imputation can directly use the mice package.

Quality Control (QC) Summary

• New & Enhanced Features:
  • Added a cor_method argument to qc_summary_data() for customizable correlation methods.
    
  • Merged qc_summary_data() and qc_summary_metadata() into a single function, hd_qc_summary(), which now visualizes all metadata columns.
• Removed Normality Checks:
  Removed automatic normality checks for large datasets due to sensitivity issues. Instead, users are encouraged to use histograms or QQ plots for this purpose.

Differential Expression Analysis

• Streamlined Functions:
  • Merged do_limma() and do_limma_continuous() into hd_de_limma(), which auto-detects variable types.
    
  • hd_plot_volcano() is now a standalone function to reduce the number of arguments but still keep the flexibility.
• Improved Customization:
  Added a user_defined_proteins argument to hd_plot_volcano(), allowing users to label specific proteins on volcano plots.

Classification Models

• Model Updates:
  Merged multiclass and binary models.

• Enhanced Visualizations:

  • Multiclassification models now include a variable importance plot.
    
  • AUC bar plots were removed to make the output more consistent with the binary classification models.
  • Probability plots were added to visualize the distribution of probabilities for each class.

Visualization Functions

• New Functions:
  • Added plot_feature_summary_heatmap(): Summarizes differential expression and classification model results in a single heatmap.
  • Added plot_feature_summary_network(): Summarizes differential expression or classification model features results in a single network.
• Bug Fixes & Improvements:
  Fixed color-matching issues in bar plots for hd_plot_de_summary() and hd_plot_model_summary() caused by frequency ties. Bars are now colored correctly.

Pathway Enrichment Analysis

• New Features
  • Added ready-to-use ORA background datasets for hd_ora().
  • Added hd_show_backgrounds(): See available backgrounds for ORA.
• Streamlined Outputs:
  Removed unnecessary returns from hd_plot_gsea().

Clustering

• New Features:
  • Added hd_cluster_samples(): Clusters samples based on selected features in k clusters. The number k is either user-defined or determined using the gap statistic.
  • Added hd_assess_clusters(): Assesses the quality of clustering using the cluster’s Jaccard index and sample size.

New Analysis Module: WGCNA

Added the Weighted Gene Co-expression Network Analysis (WGCNA) module for network analysis.

• Module Functions:
  • hd_wgcna(): Performs WGCNA analysis on the data.
  • hd_plot_wgcna(): Visualizes the WGCNA results.

---

HDAnalyzeR 1.0.0 (2024-08-19)

Initial release of HDAnalyzeR.