Extended Data Fig. 1: Construction of PanBaRT20.

The Morex and Barke GsRTDs were used as examples to illustrate the construction of PanBaRT20 from 20 GsRTDs. a, The transcripts in each GsRTD gene were collapsed into an exon union set (step 1). The union sets of all the GsRTD genes were mapped to the PSVCP pan-genome using Minimap2 (step 2). This ensured that all the transcripts from the same gene were mapped to the same genomic loci on the PSVCP pan-genome (step 3). b, The overlapped transcripts were assigned the same gene ID. The multiple-exon transcripts that shared identical intron combinations were merged, and the furthest start and end of these transcripts were taken as the transcript start site (TSS) and end site (TES) of the merged transcript (step 4). The overlapped mono-exon transcripts were merged into one transcript with the furthest starting and ending as the TSS and TES (step 5). If a set of overlapped transcripts were located entirely within the introns of other transcripts, they were assigned a separate gene ID (step 6). c, After assigning new gene and transcript IDs to the PanBaRT20 gene models, a look-up table was created to record the gene and transcript associations between PanBaRT20 and 20 GsRTDs.