Add include_version parameter to import_gencode and remove_y_par to add_gencode_transcript_annotations

[jkgoodrich]

Summary

This PR updates the GENCODE import and transcript annotation functionality to better handle transcript ID versions and PAR region duplicates.

Changes

1. Update import_gencode to keep full transcript ID including version if requested

• When include_version=True, the function now properly preserves the full gene_id and transcript_id including version numbers in separate fields (gene_id_version and transcript_id_version)
• This allows downstream functions to access both the versioned and non-versioned IDs as needed

2. Fix add_gencode_transcript_annotations to include start and end position

• Updated the default annotations parameter in add_gencode_transcript_annotations to include start_position and end_position by default
• The function now properly calculates and includes transcript start and end positions from the GENCODE interval data
• Updated documentation to reflect that these annotations are included by default

3. Add remove_y_par parameter to handle PAR region transcript ID duplicates

• Added remove_y_par parameter (default: True) to add_gencode_transcript_annotations
• This parameter filters out Y chromosome PAR (pseudoautosomal region) features to prevent duplicate transcript IDs
• PAR regions can have identical transcript IDs on both chrX and chrY, which can cause issues in constraint calculations
• The filtering uses the transcript_id_version field to identify Y_PAR transcripts (those ending with "Y_PAR")
• Updated function documentation to explain the PAR region handling and requirements

Technical Details

• The remove_y_par functionality requires that the input GENCODE table includes the transcript_id_version field, which is available when import_gencode is called with include_version=True
• The PAR region filtering prevents duplicate transcript annotations that could skew constraint calculations
• Start and end position annotations are computed from the GENCODE interval data and provide essential transcript boundary information

Impact

• Improves data quality by preventing PAR region duplicates in constraint calculations
• Provides more complete transcript annotation data including positional information
• Maintains backward compatibility while adding new functionality
• Enables better handling of GENCODE data with version information

[ch-kr]

a couple documentation suggestions and one question