Fix bubble map colorbar title and p-value column name issue

[SuhasSrinivasan]

This pull request introduces improvements to the handling and visualization of MEA enrichment results, focusing on better support for custom column names and enhanced labeling in bubble-map plots. The changes improve both the flexibility and clarity of the enrichment analysis workflow, especially for users working with MEA NES and p-value data. The update also includes new and expanded test coverage for these features.

Key changes include:

Customization and Flexibility in Bubble-map Plotting

• The plot_bubblemap function now supports optional arguments lff_cbar_title and pval_legend_title, allowing users to customize the colorbar and p-value legend titles [1] [2] [3] [4].
• Docstrings for plot_bubblemap and related parameters have been updated for clarity, reflecting the new customization options and making the documentation more accurate and user-friendly [1] [2] [3] [4].

Improved MEA Enrichment Results Combination

• The combine_mea_enrichment_results function now prefers the native MEA p-value column for unadjusted p-values, while still supporting the legacy pvalue spelling. This ensures more accurate and robust handling of MEA result tables with varying column names [1] [2] [3].

Expanded and Updated Test Coverage

• Added new tests for the MEA enrichment results combination logic, verifying correct selection of p-value columns and honoring explicit user input.
• Added tests for bubble-map plotting to ensure custom legend and colorbar titles are correctly applied and that defaults are preserved [1] [2].

Documentation

• The CHANGELOG.md has been updated to reflect the new features and bugfixes, including the ability to customize bubble-map colorbar and legend titles and improved MEA result combination logic.

Testing Environment

• The test configuration now sets the MPLBACKEND environment variable to "Agg" to ensure matplotlib tests run headlessly and deterministically.