Integrative Genomics Viewer (IGV) snapshots showing five categories of gene models optimized by transcriptomic data, including new gene (A), exon-altered gene (B), fused gene (C), partitioned gene (D) and orientation-reversed gene (E). Low-confidence genes (F) were not supported by RNA-seq data, thus retaining their annotations in draft v2 as in TGD2021. Tracks from top to bottom were RNA-seq (growth, starvation for 24 h, and conjugation at 4, 5, 6, 8 and 10 h), Nanopore DRS coverage and reads alignment, and the gene models of draft v2 and TGD2021. Reads and gene models in pink represented transcription on the sense strand, and those in purple on the antisense strand.