-Fixed operator precedence bug in count initialization to restore accurate FL-weighted transcript counts
-Fixed coordinate mismatch in build_meta_coverage_plot for ISM transcripts spanning introns to prevent valid transcripts from being dropped

…entages (html)

…d fix PDF Transcript Classification table values sorting bug

…y basic features, and fix empty DataTables in hidden dropdown tabs

… the R script subprocess to support directory paths natively containing spaces.

…and align tabs

…ce transcript body coverage by structural category plot

…utions to reads mode and change titles for clarity

… issues and package incompatibilities

…tural category to reports

… by isoform ID, using _count (FL) sum per (CB, junction_type) instead of groupby on comma-separated CB string

…rms mode

  - Fix genes_by_cb: explode CB/FL and sum FL per (CB, gene) instead of n()
  - Fix all junction-by-category blocks (PDF junc_aug, HTML junc_aug_html):
    replace broken join-by-CB+isoform with CB/FL explode + isoform-key-only join
  - Fix RT-switching junction figures (junc_rt): same CB/FL explode pattern;
    use sum(count[RTS_bool]) instead of sum(RTS_bool)
  - Fix mono-exonic figure: FL-explode cls_valid, use sum(count) instead of n()
  - Fix exon count bins by category: sum(count) instead of dplyr::n()
  - Fix exon count profile by category: sum(count) instead of dplyr::n()
  - Fix mean exons per category: FL-weighted mean (sum(exons*count)/sum(count))
    instead of unweighted median on non-exploded data
  - Replace complete(..., fill=list(...=0)) with NA fill for absent (CB, category)
    pairs across all junction/exon violin plots to eliminate zero-inflation

…edundancy

… Mean Exons violin over-smoothing

… tab switch