T2T 与前版本比较

中文

作用：把每个新 T2T 组装与同一 accession 的已发表旧组装比较（Gifu T2T vs Gifu old；MG20 T2T vs MG20 old）：MUMmer4 + SyRI 变异、由比对区间补集得到的 PAV，以及 T2T 特有新序列按注释区域的分区。 入口：GifuT2T_GifuOld/GifuT2T_vs_GifuOld_variation_finder.sh 输入：Gifu 用 Lotus_Gifu_Old.fa（seqkit 清理为 Lotus_Gifu_Old1.fa）与 Lotus_GifuT2T_v1.0.fasta，MG20 用 Lotus_MG20_Old.fa 与 Lotus_MG20T2T_v1.0.fasta，以及 genomes.txt；PAV 需对应的 <prefix>.filtered.coords；calc_bin_coverage.sh 用 gifuT2T_newseq.bed 与 {PROJ_04_LOTUS_GENOME}/00_assembly_result/GifuT2T/Gifu_100k.bed；run.sh/run.py 用 gifuT2T_newseq.bed（或 mg20T2T_newseq.bed）与 cent.bed、rDNA.bed、tel.bed、repeat2.bed。 输出：<p>.delta/.filtered.delta、dnadiff 报告、syri.out、Old_<type>.bed 与 T2T_<type>.bed（SNP、INS、DEL、INV、TRANS、INVTR、DUP、INVDP、PAV）、stdout 计数表与 *_findSV.pdf；PAV 出 .genome、<asm>.aligned.bed、<asm>.PAV.bed、PAV.summary；calc_bin_coverage.sh 出 gifu_100kb_coverage.tsv；run.sh/run.py 出各区域 bp 与占比及 new_sequences_region_proportions.pdf。 运行：每个比较各自一个工作目录，先跑 variation finder 再跑 PAV：cd GifuT2T_GifuOld && bash GifuT2T_vs_GifuOld_variation_finder.sh && bash GifuT2T_vs_GifuOld_get_PAV_from_coords.sh（MG20 同理）；随后 bash calc_bin_coverage.sh，并在 Gifu_new/、MG20_new/ 中各跑 bash run.sh 或 python3 run.py。 工具：MUMmer4 4.0.1（nucmer -c 1000 --mum --maxgap=1000、delta-filter -i 90 -l 1000 -q、show-coords -THrd、dnadiff -d -p）、SyRI 1.7.1、plotsr 1.1.1、samtools 1.21、bedtools 2.31.1（merge -d 50、complement -g、intersect、subtract）、seqkit 2.10.1（grep -v -r -p）、GNU Awk 5.1.0、Python 3 + pandas/matplotlib、CSUB（-q c01、-n 64）。

English

**Purpose**: Compare each new T2T assembly with the published earlier assembly of the same accession (Gifu T2T vs Gifu old; MG20 T2T vs MG20 old): MUMmer4 + SyRI variants, PAV from the complement of the aligned intervals, and a partition of the T2T-specific new sequences by the annotated regions they fall in. **Entry point**: `GifuT2T_GifuOld/GifuT2T_vs_GifuOld_variation_finder.sh` **Inputs**: for Gifu, `Lotus_Gifu_Old.fa` (cleaned by seqkit to `Lotus_Gifu_Old1.fa`) and `Lotus_GifuT2T_v1.0.fasta`; for MG20, `Lotus_MG20_Old.fa` and `Lotus_MG20T2T_v1.0.fasta`; plus `genomes.txt`; the PAV step needs the matching `.filtered.coords`; `calc_bin_coverage.sh` uses `gifuT2T_newseq.bed` and `{PROJ_04_LOTUS_GENOME}/00_assembly_result/GifuT2T/Gifu_100k.bed`; `run.sh`/`run.py` use `gifuT2T_newseq.bed` (or `mg20T2T_newseq.bed`) with `cent.bed`, `rDNA.bed`, `tel.bed` and `repeat2.bed`. **Outputs**: `

.delta`/`.filtered.delta`, dnadiff reports, `syri.out`, `Old_.bed` and `T2T_.bed` (`type` = SNP, INS, DEL, INV, TRANS, INVTR, DUP, INVDP, PAV), a count table on stdout and `*_findSV.pdf`; the PAV step writes `.genome`, `.aligned.bed`, `.PAV.bed` and `PAV.summary`; `calc_bin_coverage.sh` writes `gifu_100kb_coverage.tsv`; `run.sh`/`run.py` print per-region bp and ratios and write `new_sequences_region_proportions.pdf`. **Run**: give each comparison its own working directory and run the variation finder before the PAV step: `cd GifuT2T_GifuOld && bash GifuT2T_vs_GifuOld_variation_finder.sh && bash GifuT2T_vs_GifuOld_get_PAV_from_coords.sh` (the same for MG20); then `bash calc_bin_coverage.sh`, and `bash run.sh` or `python3 run.py` in `Gifu_new/` and `MG20_new/`. **Tools**: MUMmer4 4.0.1 (`nucmer -c 1000 --mum --maxgap=1000`, `delta-filter -i 90 -l 1000 -q`, `show-coords -THrd`, `dnadiff -d -p`), SyRI 1.7.1, plotsr 1.1.1, samtools 1.21, bedtools 2.31.1 (`merge -d 50`, `complement -g`, `intersect`, `subtract`), seqkit 2.10.1 (`grep -v -r -p`), GNU Awk 5.1.0, Python 3 with pandas and matplotlib, CSUB (`-q c01`, `-n 64`).