Supplementary Methods (Additional file 2)
Reproducible command-line workflow to quantify organelle-derived reads

Overview
This workflow estimates the fraction of reads that map to either the chloroplast or mitochondrial genome (union), and splits organelle-like reads into:
- cpDNA only: reads mapping to the chloroplast reference but not to the mitochondrial reference
- mtDNA only: reads mapping to the mitochondrial reference but not to the chloroplast reference
- cpDNA and mtDNA: reads mapping to both references (intersection)
- cpDNA or mtDNA: reads mapping to either reference (union)

Requirements
- SRA Toolkit (prefetch, fasterq-dump) OR enaDataGet/asaspera as an alternative
- BWA (bwa mem)
- SAMtools
- coreutils (awk, sort, comm)
- GNU parallel (optional)

Inputs
- List of run accessions (one per line), e.g., runs.txt
- Reference FASTA files:
  - cp.fa (Silene latifolia chloroplast genome; GenBank NC_016730)
  - mt.fa (Silene latifolia mitochondrial genome; GenBank NC_014487)

1) Prepare references
bwa index cp.fa
bwa index mt.fa

2) Download reads (example for one run)
# Using SRA Toolkit (paired-end runs will produce *_1.fastq and *_2.fastq)
prefetch SRR2050365
fasterq-dump --split-files --threads 8 SRR2050365 -O fastq/

3) Map reads to organelle references
# Single-end:
bwa mem -t 8 cp.fa fastq/SRR2050365.fastq | samtools sort -@ 4 -o bam/SRR2050365.cp.bam
bwa mem -t 8 mt.fa fastq/SRR2050365.fastq | samtools sort -@ 4 -o bam/SRR2050365.mt.bam

# Paired-end:
bwa mem -t 8 cp.fa fastq/SRR2050365_1.fastq fastq/SRR2050365_2.fastq | samtools sort -@ 4 -o bam/SRR2050365.cp.bam
bwa mem -t 8 mt.fa fastq/SRR2050365_1.fastq fastq/SRR2050365_2.fastq | samtools sort -@ 4 -o bam/SRR2050365.mt.bam

samtools index bam/SRR2050365.cp.bam
samtools index bam/SRR2050365.mt.bam

4) Count total reads (N_total)
# For FASTQ:
# Single-end:
N_total=$(awk 'END{print NR/4}' fastq/SRR2050365.fastq)
# Paired-end (count read ends, consistent with SRA "reads"):
N_total=$(($(awk 'END{print NR/4}' fastq/SRR2050365_1.fastq)+$(awk 'END{print NR/4}' fastq/SRR2050365_2.fastq)))

5) Extract mapped read IDs (flag != 4) and compute set operations
# Mapped to chloroplast:
samtools view -F 4 bam/SRR2050365.cp.bam | awk '{print $1}' | sort -u > ids/SRR2050365.cp.ids
# Mapped to mitochondrion:
samtools view -F 4 bam/SRR2050365.mt.bam | awk '{print $1}' | sort -u > ids/SRR2050365.mt.ids

# Intersection (cpDNA and mtDNA):
comm -12 ids/SRR2050365.cp.ids ids/SRR2050365.mt.ids > ids/SRR2050365.both.ids
# cpDNA only:
comm -23 ids/SRR2050365.cp.ids ids/SRR2050365.mt.ids > ids/SRR2050365.cp_only.ids
# mtDNA only:
comm -13 ids/SRR2050365.cp.ids ids/SRR2050365.mt.ids > ids/SRR2050365.mt_only.ids
# Union (cpDNA or mtDNA):
cat ids/SRR2050365.cp.ids ids/SRR2050365.mt.ids | sort -u > ids/SRR2050365.cp_or_mt.ids

N_cp_only=$(wc -l < ids/SRR2050365.cp_only.ids)
N_mt_only=$(wc -l < ids/SRR2050365.mt_only.ids)
N_both=$(wc -l < ids/SRR2050365.both.ids)
N_cp_or_mt=$(wc -l < ids/SRR2050365.cp_or_mt.ids)

6) Compute organelle-derived fraction
Organelle_percent=$(python3 - <<PY
N_total=int(${N_total})
N_cp_or_mt=int(${N_cp_or_mt})
print(round(100.0*N_cp_or_mt/N_total, 1))
PY
)

7) Report one-line summary
echo -e "SRR2050365\t${N_total}\t${N_cp_only}\t${N_mt_only}\t${N_both}\t${N_cp_or_mt}\t${Organelle_percent}"

Notes
- If you prefer stricter mapping, add a mapping-quality threshold (e.g., samtools view -q 20). This will reduce spurious alignments but may change the values relative to Table S1.
- The estimate is "organelle-like": reads from nuclear insertions of organelle DNA (NUMTs/NUPTs) may also map to organelle references.