rDNA 拷贝数
作用:从测序深度估计 Gifu 与 MG20 的 45S 与 5S rDNA 拷贝数;把 T2T 装配的 rDNA 阵列屏蔽后追加一个 45S 与一个 5S 代表重复构成折叠参考,WGS/HiFi reads 比对后用 mosdepth 量深度,组件深度除以单拷贝 BUSCO 位点深度中位数即得拷贝数,并与装配来源计数并列。
入口:run_rDNA_CN.sh
输入:装配 {PROJ_04_LOTUS_GENOME}/00_assembly_result/{GifuT2T/Lotus_GifuT2T_v1.0.fasta,MG20T2T/Lotus_MG20T2T_v1.0.fasta};WGS 01_data/<acc>/WGS/<acc>_Clean_R1.fq 与 _R2.fq;HiFi 01_data/<acc>/hifi/<acc>_hifi.fastq.gz;BUSCO 02_assembly/<acc>/08.quality/02.busco/<acc>_lotus/run_embryophyta_odb10/full_table.tsv;rDNA 阵列 BED 03_annotation/<acc>/rDNA/rRNA.bed。
输出:${OUTDIR} = {PROJ_04_LOTUS_GENOME}/03_annotation/rDNA_CN,每材料一个子目录:02_rdna/(rDNA_arrays.bed、代表重复 .fa/.gff)、03_busco/(*_BUSCO_singlecopy.bed)、04_ref/(*_CN_reference.fa 及索引)、05_map/(*_{WGS,HiFi}_CN.bam)、06_depth/(mosdepth *.regions.bed.gz)、07_cn/(*_{WGS,HiFi}_CN_summary.tsv),另有合并表 final_rDNA_CN.tsv。
运行:bash run_rDNA_CN.sh(#CSUB -J rDNA_CN、队列 c01、-n 64、span[hosts=1];一次提交内先 Gifu 后 MG20)。
工具:bwa mem -t 50;minimap2 -ax map-hifi -t 50;samtools faidx、sort -@ 16、index、flagstat;bedtools maskfasta、intersect -v;mosdepth -t 8 --by <bed>;Barrnap --kingdom euk 校验代表重复;Python 3(pandas、numpy)。
English
Purpose: Estimate the 45S and 5S rDNA copy number of Gifu and MG20 from read depth; the rDNA arrays of the T2T assembly are masked and one representative 45S and one representative 5S repeat are appended to form a collapsed reference, WGS and HiFi reads are mapped and depth is measured with mosdepth, and each rDNA component’s depth divided by the median depth over single-copy BUSCO loci gives the copy number, set beside the assembly-derived counts.
Entry point: run_rDNA_CN.sh
Inputs: the assemblies {PROJ_04_LOTUS_GENOME}/00_assembly_result/{GifuT2T/Lotus_GifuT2T_v1.0.fasta,MG20T2T/Lotus_MG20T2T_v1.0.fasta}; WGS 01_data/<acc>/WGS/<acc>_Clean_R1.fq and _R2.fq; HiFi 01_data/<acc>/hifi/<acc>_hifi.fastq.gz; BUSCO 02_assembly/<acc>/08.quality/02.busco/<acc>_lotus/run_embryophyta_odb10/full_table.tsv; the rDNA array BED 03_annotation/<acc>/rDNA/rRNA.bed.
Outputs: ${OUTDIR} = {PROJ_04_LOTUS_GENOME}/03_annotation/rDNA_CN, one subdirectory per accession: 02_rdna/ (rDNA_arrays.bed, the representative repeats .fa/.gff), 03_busco/ (*_BUSCO_singlecopy.bed), 04_ref/ (*_CN_reference.fa and indexes), 05_map/ (*_{WGS,HiFi}_CN.bam), 06_depth/ (mosdepth *.regions.bed.gz), 07_cn/ (*_{WGS,HiFi}_CN_summary.tsv), plus the merged table final_rDNA_CN.tsv.
Run: bash run_rDNA_CN.sh (#CSUB -J rDNA_CN, queue c01, -n 64, span[hosts=1]; Gifu then MG20 in one submission).
Tools: bwa mem -t 50; minimap2 -ax map-hifi -t 50; samtools faidx, sort -@ 16, index, flagstat; bedtools maskfasta, intersect -v; mosdepth -t 8 --by <bed>; Barrnap --kingdom euk to verify the representative repeats; Python 3 (pandas, numpy).
以下为该步骤目录中的实际文件(同步快照,与该目录内容一致)。
run_rDNA_CN.sh · Shell · 354 行
#!/bin/bash#CSUB -J rDNA_CN#CSUB -q c01#CSUB -o rDNA_CN.out#CSUB -e rDNA_CN.error#CSUB -n 64#CSUB -R span[hosts=1]
# ==========================================================================# rDNA copy-number estimation pipeline# Coverage-based CNV (Illumina WGS) + HiFi independent validation# For Gifu and MG20 T2T assemblies# ==========================================================================
set -eo pipefail
THREADS=50 # max threads for mappingMOSDEPTH_THREADS=8PROJ={PROJ_04_LOTUS_GENOME}OUTDIR=${PROJ}/03_annotation/rDNA_CNmkdir -p ${OUTDIR}
# ── Paths ──────────────────────────────────────────────────────────────GIFU_FA=${PROJ}/00_assembly_result/GifuT2T/Lotus_GifuT2T_v1.0.fastaMG20_FA=${PROJ}/00_assembly_result/MG20T2T/Lotus_MG20T2T_v1.0.fasta
GIFU_WGS_R1=${PROJ}/01_data/Gifu/WGS/Gifu_Clean_R1.fqGIFU_WGS_R2=${PROJ}/01_data/Gifu/WGS/Gifu_Clean_R2.fqMG20_WGS_R1=${PROJ}/01_data/MG20/WGS/MG20_Clean_R1.fqMG20_WGS_R2=${PROJ}/01_data/MG20/WGS/MG20_Clean_R2.fq
GIFU_HIFI=${PROJ}/01_data/Gifu/hifi/Gifu_hifi.fastq.gzMG20_HIFI=${PROJ}/01_data/MG20/hifi/MG20_hifi.fastq.gz
GIFU_BUSCO=${PROJ}/02_assembly/Gifu/08.quality/02.busco/Gifu_lotus/run_embryophyta_odb10/full_table.tsvMG20_BUSCO=${PROJ}/02_assembly/MG20/08.quality/02.busco/MG20_lotus/run_embryophyta_odb10/full_table.tsv
GIFU_RDNA_BED=${PROJ}/03_annotation/Gifu/rDNA/rRNA.bedMG20_RDNA_BED=${PROJ}/03_annotation/MG20/rDNA/rRNA.bed
# ── Representative repeat coordinates (1-based, extracted from T2T) ────# Gifu Chr2 45S: first complete unit (~10 kb, from first 5.8S to before second 5.8S)GIFU_45S_CHR=Chr2GIFU_45S_START=537436GIFU_45S_END=547437# Gifu Chr2 5S: first 5S cluster (~300 bp)GIFU_5S_CHR=Chr2GIFU_5S_START=24476202GIFU_5S_END=24476511
# MG20 Chr2 45S: first complete unit (~13 kb)MG20_45S_CHR=Chr2MG20_45S_START=1528120MG20_45S_END=1541119# MG20 Chr2 5S: first 5S cluster (~300 bp)MG20_5S_CHR=Chr2MG20_5S_START=26673363MG20_5S_END=26673672
# ── rDNA array regions for masking (BED) ───────────────────────────────make_rdna_bed() { local out=$1 species=$2 case $species in Gifu) cat > ${out} <<'BEDEOF'Chr2 537436 949316 45S_NOR_IChr2 24476202 25027540 5S_arrayChr5 14943604 15344676 45S_Chr5Chr6 8844950 8943185 45S_Chr6_IChr6 26125598 27223832 45S_Chr6_IIBEDEOF ;; MG20) cat > ${out} <<'BEDEOF'Chr2 1528119 2857257 45S_NOR_IChr2 26673363 27897859 5S_arrayChr5 15013664 15659889 45S_Chr5Chr6 15057396 15451961 45S_Chr6_IChr6 26255619 26622326 45S_Chr6_IIBEDEOF ;; esac echo " rDNA arrays BED -> ${out}"}
# ==========================================================================# process_one_species <name> <fa> <wgs_R1> <wgs_R2> <hifi> <busco_tsv># ==========================================================================process_one_species() { SP_NAME=$1 SP_FA=$2 SP_WGS_R1=$3 SP_WGS_R2=$4 SP_HIFI=$5 SP_BUSCO=$6 SP_C45=$7 SP_C45s=$8 SP_C45e=$9 SP_C5=${10} SP_C5s=${11} SP_C5e=${12}
echo "" echo "================================================================" echo " Processing ${SP_NAME}" echo "================================================================"
WD=${OUTDIR}/${SP_NAME} mkdir -p ${WD}/02_rdna ${WD}/03_busco ${WD}/04_ref ${WD}/05_map ${WD}/06_depth ${WD}/07_cn
# ── Step 1: rDNA arrays BED ───────────────────────────────────── echo "[${SP_NAME}] Step 1: rDNA arrays BED" make_rdna_bed ${WD}/02_rdna/rDNA_arrays.bed ${SP_NAME}
# ── Step 2: Extract representative 45S & 5S repeats ────────────── echo "[${SP_NAME}] Step 2: Extract representative repeats" echo " Extracting 45S: ${SP_C45}:${SP_C45s}-${SP_C45e}" echo " Extracting 5S: ${SP_C5}:${SP_C5s}-${SP_C5e}" samtools faidx ${SP_FA} "${SP_C45}:${SP_C45s}-${SP_C45e}" \ | sed "s/^>.*/>${SP_NAME}_45S_repeat/" \ > ${WD}/02_rdna/${SP_NAME}_45S_repeat.fa samtools faidx ${SP_FA} "${SP_C5}:${SP_C5s}-${SP_C5e}" \ | sed "s/^>.*/>${SP_NAME}_5S_repeat/" \ > ${WD}/02_rdna/${SP_NAME}_5S_repeat.fa
# Verify with barrnap barrnap --kingdom euk ${WD}/02_rdna/${SP_NAME}_45S_repeat.fa \ > ${WD}/02_rdna/${SP_NAME}_45S_repeat.gff 2>/dev/null barrnap --kingdom euk ${WD}/02_rdna/${SP_NAME}_5S_repeat.fa \ > ${WD}/02_rdna/${SP_NAME}_5S_repeat.gff 2>/dev/null echo " 45S repeat: $(grep -c 'rRNA' ${WD}/02_rdna/${SP_NAME}_45S_repeat.gff || echo 0) rRNA hits" echo " 5S repeat: $(grep -c 'rRNA' ${WD}/02_rdna/${SP_NAME}_5S_repeat.gff || echo 0) rRNA hits"
# ── Step 3: Build collapsed rDNA reference ─────────────────────── echo "[${SP_NAME}] Step 3: Build collapsed reference" bedtools maskfasta \ -fi ${SP_FA} \ -bed ${WD}/02_rdna/rDNA_arrays.bed \ -fo ${WD}/04_ref/${SP_NAME}_T2T.rDNA_masked.fa cat ${WD}/04_ref/${SP_NAME}_T2T.rDNA_masked.fa \ ${WD}/02_rdna/${SP_NAME}_45S_repeat.fa \ ${WD}/02_rdna/${SP_NAME}_5S_repeat.fa \ > ${WD}/04_ref/${SP_NAME}_CN_reference.fa samtools faidx ${WD}/04_ref/${SP_NAME}_CN_reference.fa bwa index ${WD}/04_ref/${SP_NAME}_CN_reference.fa 2> ${WD}/04_ref/bwa_index.log echo " Reference: $(wc -c < ${WD}/04_ref/${SP_NAME}_CN_reference.fa) bp"
# ── Step 4: BUSCO single-copy baseline ─────────────────────────── echo "[${SP_NAME}] Step 4: BUSCO single-copy baseline" awk -F'\t' ' BEGIN{OFS="\t"} !/^#/ && $2=="Complete" { s=$4; e=$5; if(s>e){tmp=s; s=e; e=tmp} print $3, s-1, e, $1 }' ${SP_BUSCO} > ${WD}/03_busco/${SP_NAME}_BUSCO_singlecopy.raw.bed
bedtools intersect -v \ -a ${WD}/03_busco/${SP_NAME}_BUSCO_singlecopy.raw.bed \ -b ${WD}/02_rdna/rDNA_arrays.bed \ > ${WD}/03_busco/${SP_NAME}_BUSCO_singlecopy.bed
BUSCO_N=$(wc -l < ${WD}/03_busco/${SP_NAME}_BUSCO_singlecopy.bed) echo " Single-copy BUSCO loci: ${BUSCO_N}"
# ── Step 5: Illumina WGS mapping ───────────────────────────────── echo "[${SP_NAME}] Step 5: Illumina WGS mapping (bwa mem, ${THREADS} threads)" bwa mem -t ${THREADS} -R "@RG\tID:${SP_NAME}\tSM:${SP_NAME}" \ ${WD}/04_ref/${SP_NAME}_CN_reference.fa \ ${SP_WGS_R1} ${SP_WGS_R2} \ | samtools sort -@ 16 -o ${WD}/05_map/${SP_NAME}_WGS_CN.bam - samtools index ${WD}/05_map/${SP_NAME}_WGS_CN.bam samtools flagstat ${WD}/05_map/${SP_NAME}_WGS_CN.bam \ > ${WD}/05_map/${SP_NAME}_WGS_CN.flagstat.txt echo " WGS BAM done"
# ── Step 6: rDNA component BED from representative repeats ─────── echo "[${SP_NAME}] Step 6: rDNA component BED" { grep 'rRNA' ${WD}/02_rdna/${SP_NAME}_45S_repeat.gff | \ awk -F'\t' '{ type="" if($9 ~ /5_8S|5\.8S/) type="5.8S" else if($9 ~ /18S/) type="18S" else if($9 ~ /28S/) type="28S" if(type) print $1"\t"$4-1"\t"$5"\t"type }' grep 'rRNA' ${WD}/02_rdna/${SP_NAME}_5S_repeat.gff | \ awk -F'\t' '{ type="" if($9 ~ /5S/ && $9 !~ /5_8S|5\.8S/) type="5S" if(type) print $1"\t"$4-1"\t"$5"\t"type }' } > ${WD}/06_depth/${SP_NAME}_rDNA_components.bed
N_comp=$(wc -l < ${WD}/06_depth/${SP_NAME}_rDNA_components.bed) echo " rDNA components: ${N_comp} regions"
# ── Step 7: mosdepth for rDNA + BUSCO (WGS) ───────────────────── echo "[${SP_NAME}] Step 7: mosdepth (WGS)" mosdepth -t ${MOSDEPTH_THREADS} --by ${WD}/06_depth/${SP_NAME}_rDNA_components.bed \ ${WD}/06_depth/${SP_NAME}_WGS_rDNA \ ${WD}/05_map/${SP_NAME}_WGS_CN.bam 2>/dev/null mosdepth -t ${MOSDEPTH_THREADS} --by ${WD}/03_busco/${SP_NAME}_BUSCO_singlecopy.bed \ ${WD}/06_depth/${SP_NAME}_WGS_BUSCO \ ${WD}/05_map/${SP_NAME}_WGS_CN.bam 2>/dev/null echo " mosdepth done"
# ── Step 8: Calculate CN (WGS) ─────────────────────────────────── echo "[${SP_NAME}] Step 8: Calculate CN (WGS)" python3 - "${SP_NAME}" "${WD}" "WGS" <<'PYEOF'import sys, gzip, numpy as np, pandas as pd
def read_mosdepth(path): opener = gzip.open if path.endswith('.gz') else open rows = [] with opener(path, 'rt') as f: for line in f: x = line.rstrip().split('\t') rows.append({'chrom':x[0],'start':int(x[1]),'end':int(x[2]), 'name':x[3] if len(x)>=5 else '.','depth':float(x[-1])}) return pd.DataFrame(rows)
name, wd, tag = sys.argv[1], sys.argv[2], sys.argv[3]busco = read_mosdepth(f'{wd}/06_depth/{name}_{tag}_BUSCO.regions.bed.gz')rdna = read_mosdepth(f'{wd}/06_depth/{name}_{tag}_rDNA.regions.bed.gz')D_single = np.median(busco['depth'])
print(f" D_single (BUSCO median) = {D_single:.3f}x")for _,r in rdna.iterrows(): cn = r['depth'] / D_single if D_single>0 else 0 print(f" {r['name']:6s} depth={r['depth']:9.3f} CN={cn:.2f}")
sub = rdna[rdna['name'].isin(['18S','5.8S','28S'])]cn_45s = np.median(sub['depth']) / D_single if D_single>0 and len(sub)>0 else np.nansub5 = rdna[rdna['name']=='5S']cn_5s = np.median(sub5['depth'])/D_single if D_single>0 and len(sub5)>0 else np.nan
with open(f'{wd}/07_cn/{name}_{tag}_CN_summary.tsv','w') as f: f.write("accession\ttype\testimated_CN\tsingle_copy_depth\n") f.write(f"{name}\t45S\t{cn_45s:.3f}\t{D_single:.3f}\n") f.write(f"{name}\t5S\t{cn_5s:.3f}\t{D_single:.3f}\n")print(f" => {name} {tag}: 45S={cn_45s:.2f} 5S={cn_5s:.2f}")PYEOF
# ── Step 9: HiFi mapping (independent validation) ──────────────── echo "[${SP_NAME}] Step 9: HiFi mapping (minimap2, ${THREADS} threads)" minimap2 -ax map-hifi -t ${THREADS} -R "@RG\tID:${SP_NAME}_hifi\tSM:${SP_NAME}" \ ${WD}/04_ref/${SP_NAME}_CN_reference.fa ${SP_HIFI} \ | samtools sort -@ 16 -o ${WD}/05_map/${SP_NAME}_HiFi_CN.bam - samtools index ${WD}/05_map/${SP_NAME}_HiFi_CN.bam echo " HiFi BAM done"
# ── Step 10: mosdepth + CN for HiFi ───────────────────────────── echo "[${SP_NAME}] Step 10: CN (HiFi)" mosdepth -t ${MOSDEPTH_THREADS} --by ${WD}/06_depth/${SP_NAME}_rDNA_components.bed \ ${WD}/06_depth/${SP_NAME}_HiFi_rDNA \ ${WD}/05_map/${SP_NAME}_HiFi_CN.bam 2>/dev/null mosdepth -t ${MOSDEPTH_THREADS} --by ${WD}/03_busco/${SP_NAME}_BUSCO_singlecopy.bed \ ${WD}/06_depth/${SP_NAME}_HiFi_BUSCO \ ${WD}/05_map/${SP_NAME}_HiFi_CN.bam 2>/dev/null
python3 - "${SP_NAME}" "${WD}" "HiFi" <<'PYEOF'import sys, gzip, numpy as np, pandas as pddef read_mosdepth(path): opener = gzip.open if path.endswith('.gz') else open rows = [] with opener(path, 'rt') as f: for line in f: x = line.rstrip().split('\t') rows.append({'chrom':x[0],'start':int(x[1]),'end':int(x[2]), 'name':x[3] if len(x)>=5 else '.','depth':float(x[-1])}) return pd.DataFrame(rows)name, wd, tag = sys.argv[1], sys.argv[2], sys.argv[3]busco = read_mosdepth(f'{wd}/06_depth/{name}_{tag}_BUSCO.regions.bed.gz')rdna = read_mosdepth(f'{wd}/06_depth/{name}_{tag}_rDNA.regions.bed.gz')D_single = np.median(busco['depth'])sub = rdna[rdna['name'].isin(['18S','5.8S','28S'])]cn_45s = np.median(sub['depth'])/D_single if D_single>0 and len(sub)>0 else np.nansub5 = rdna[rdna['name']=='5S']cn_5s = np.median(sub5['depth'])/D_single if D_single>0 and len(sub5)>0 else np.nanwith open(f'{wd}/07_cn/{name}_{tag}_CN_summary.tsv','w') as f: f.write("accession\ttype\testimated_CN\tsingle_copy_depth\n") f.write(f"{name}\t45S\t{cn_45s:.3f}\t{D_single:.3f}\n") f.write(f"{name}\t5S\t{cn_5s:.3f}\t{D_single:.3f}\n")print(f" => {name} {tag}: 45S={cn_45s:.2f} 5S={cn_5s:.2f}")PYEOF
echo " ${SP_NAME} done."}
# ==========================================================================# Main# ==========================================================================echo "========================================"echo " rDNA copy-number estimation pipeline"echo " started at $(date)"echo "========================================"
# ── Gifu ──────────────────────────────────────────────────────────────process_one_species "Gifu" \ "${GIFU_FA}" "${GIFU_WGS_R1}" "${GIFU_WGS_R2}" "${GIFU_HIFI}" "${GIFU_BUSCO}" \ "${GIFU_45S_CHR}" ${GIFU_45S_START} ${GIFU_45S_END} \ "${GIFU_5S_CHR}" ${GIFU_5S_START} ${GIFU_5S_END}
# ── MG20 ──────────────────────────────────────────────────────────────process_one_species "MG20" \ "${MG20_FA}" "${MG20_WGS_R1}" "${MG20_WGS_R2}" "${MG20_HIFI}" "${MG20_BUSCO}" \ "${MG20_45S_CHR}" ${MG20_45S_START} ${MG20_45S_END} \ "${MG20_5S_CHR}" ${MG20_5S_START} ${MG20_5S_END}
# ==========================================================================# Final summary table# ==========================================================================echo ""echo "========================================"echo " Final summary"echo "========================================"
SUMMARY=${OUTDIR}/final_rDNA_CN.tsvcat > ${SUMMARY} <<'HEADER'accession type method assembly_CN coverage_CN single_copy_depthHEADER
for sp in Gifu MG20; do for typ in 45S 5S; do # Assembly CN from previous run (rDNA_array_summary.tsv) case "${sp}_${typ}" in Gifu_45S) asm_cn="40+16+10+76=142" ;; Gifu_5S) asm_cn="1480" ;; MG20_45S) asm_cn="129+19+38+20=206" ;; MG20_5S) asm_cn="3491" ;; esac for tag in WGS HiFi; do f=${OUTDIR}/${sp}/07_cn/${sp}_${tag}_CN_summary.tsv if [ -f "${f}" ]; then val=$(awk -F'\t' -v t="${typ}" '$2==t{print $3}' ${f}) d=$(awk -F'\t' -v t="${typ}" '$2==t{print $4}' ${f}) echo "${sp} ${typ} ${tag} ${asm_cn} ${val} ${d}" >> ${SUMMARY} fi done donedone
echo ""echo "Final table: ${SUMMARY}"cat ${SUMMARY}echo ""echo "========================================"echo " Pipeline finished at $(date)"echo "========================================"