Wgsim

# 2) Simulate 150 bp reads with a 350 bp insert size
wgsim \
  -1 150 \
  -2 150 \
  -d 350 \
  -s 30 \
  ref.fa \
  sim_R1.fq \
  sim_R2.fq

# 3) Increase mutation and indel rates for a tougher benchmark
wgsim \
  -r 0.005 \
  -R 0.20 \
  -X 0.50 \
  -e 0.01 \
  ref.fa \
  sim_R1.fq \
  sim_R2.fq

Recommended Workflow

1. Choose the reference FASTA and decide how many read pairs, what read lengths, and what insert-size distribution best match the target assay.
2. Set mutation (-r), indel (-R, -X), and sequencing error (-e) parameters explicitly instead of relying on remembered defaults.
3. Fix the random seed with -S whenever you need reproducible benchmarks.
4. Verify the emitted FASTQ pair count and use the simulated reads to evaluate alignment, calling, or QC behavior downstream.

Guardrails

• Input must be a valid FASTA reference file.
• wgsim requires two FASTQ output paths; it is a paired-end simulator.
• -h enables haplotype mode, it is not a help flag.
• GNU-style --help / --version print usage with invalid-option noise; use references/help.md or a no-argument invocation to inspect usage instead.
• Reads with too many ambiguous bases are discarded according to -A, so reference composition can affect realized output yield.