radigest

Fast in-silico restriction digest for genomics. Give it a reference FASTA (plain or .gz) or synthesize one on the fly; pass one or two enzymes; it scans, size-selects, and exports fragments as GFF3 for GBS/ddRAD, probe design, or visualization. Output order is deterministic even with multithreading.

---

Features

• Single or double digest. Double-digest keeps adjacent AB/BA by default; enable AA/BB too with -allow-same. Single-digest uses consecutive A cuts.
• IUPAC & cut offsets. Sites accept degenerate codes; the cut index comes from ^ in the site (or mid-site if missing). -strict-cuts makes missing carets an error.
• Robust FASTA I/O. Read from a path or - (STDIN), auto-detect .gz, normalize case, and trim CRLF. N in the reference does not match any site.
• Synthetic genomes. Generate a single-chromosome genome named chr1 with -sim-len, -sim-gc, -sim-seed and digest it directly—no FASTA on disk needed.
• Clean outputs. GFF3 with ID=<chr>_<n>;Length=<bp>; optional JSON summary of counts/bases per chromosome. Coordinates are 1-based closed in GFF (internally 0-based half-open).

---

Quick start

# Single digest (EcoRI) → GFF file
radigest -fasta ref.fa -enzymes EcoRI -gff fragments.gff3

# Double digest with size selection + JSON summary
radigest -fasta ref.fa -enzymes EcoRI,MseI -min 100 -max 800 -gff fragments.gff3 -json run.json

# Double digest but ALSO keep AA/BB neighbors
radigest -fasta ref.fa -enzymes EcoRI,MseI -allow-same -gff fragments.gff3

# Simulate a 10 Mb genome at 42% GC and digest
radigest -sim-len 10000000 -sim-gc 0.42 -sim-seed 123 -enzymes EcoRI,MseI -gff out.gff3

---

CLI (most used)

• -fasta <path|-> — reference FASTA; - = STDIN; .gz auto-detected.
• -enzymes E1[,E2] — one (A) or two (A,B). In double-digest, AB/BA by default.
• -min/-max — keep fragments in [min,max] bp (default min=1).
• -gff <path|-> — GFF3 out; - = STDOUT.
• -json <path> — write a run summary (counts, bases, per-chrom stats).
• -threads <n>, -v, -version, -list-enzymes.
• Simulation: -sim-len <bp>, -sim-gc <0..1>, -sim-seed <int> (emits a single chr1).
• Modes: -allow-same (keep AA/BB in double-digest), -strict-cuts (error if a site lacks ^ and would otherwise fall back to mid-site).

---

Outputs

GFF3

##gff-version 3
chr1	radigest	fragment	<start>	<end>	.	+	.	ID=chr1_1;Length=123

start/end are 1-based closed; Length is end - start + 1. Ordering is deterministic per chromosome.

JSON (optional)

{
  "enzymes": ["EcoRI","MseI"],
  "min_length": 100,
  "max_length": 800,
  "total_fragments": 123456,
  "total_bases": 7891011,
  "per_chromosome": {"chr1": {"fragments": 23456, "bases": 3456789}}
}