Restriction digests: from cut sites to fragments and a gel

Guide · Updated 2026-11-10

How restriction enzymes cut DNA at short sites, how to count the pieces from one or more enzymes in a plasmid, and how to predict the gel bands.

Open the Restriction digest simulator Free, in your browser, nothing uploaded.

Before you pour a gel, it helps to know how many pieces an enzyme will make from your DNA, how big they are and whether they will separate. The restriction digest simulator simulator works this out in your browser: paste a plasmid, choose enzymes, and it lists every cut, every fragment and a virtual gel. Check the sequence first with DNA sequence tools, and after the digest size the real bands with the gel band size estimator and join the pieces with ligation calculator.

What it is

A restriction enzyme cuts DNA only where it finds a short recognition site. EcoRI, for example, recognises GAATTC. The notation G^AATTC means the enzyme cuts the top strand after the G. Most sites read the same on both strands, so the tool searches both.

Positions are “after base”: a cut at 600 falls between bases 600 and 601 of the strand you pasted.

Counting fragments depends on the shape of the DNA:

• Linear DNA with n cuts gives n + 1 pieces.
• Circular DNA (a plasmid) with n cuts gives n pieces. Each piece is the gap to the next cut, and the last piece wraps round the origin.

For a circular molecule of length L with cuts at positions p₁ < p₂ < … < pₙ:

fragment between cut i and cut i + 1 = pᵢ₊₁ − pᵢ, and the wrap-around piece = L − pₙ + p₁

The two lengths always add up to L, which is a useful check.

How to use it

1. Sequence. Paste DNA or open a FASTA file, and choose Circular (plasmid) or Linear.
2. Enzymes. Tick any of the 12 classic enzymes, or type your own, one per line, as a name, a space and the site with ^ where the top strand is cut, for example MyEnz GT^MKAC. If the two strands are cut at different places, add _ where the bottom strand is cut.
3. Gel. Type your ladder’s band sizes in bp, and choose Each enzyme alone, then all together or All together only.
4. Read the summary. It gives the fragment sizes for the combined digest. The table lists every fragment with its start, its end and the enzyme that made each end.
5. Keep the numbers. Use Download fragments (CSV) or the report downloads for your lab book.

A worked example

Take a made-up 2,400 bp circular plasmid. EcoRI cuts it after bases 600 and 1,900. BamHI cuts after 1,000 and HindIII after 2,200.

• EcoRI alone: two cuts, so two pieces. From 600 to 1,900 is 1,900 − 600 = 1,300 bp. The wrap-around piece is 2,400 − 1,900 + 600 = 1,100 bp.
• BamHI alone or HindIII alone: one cut, so the plasmid opens into one linear piece of 2,400 bp. A single cutter like this is a common choice for opening a vector before you insert a gene.
• EcoRI and BamHI together: cuts at 600, 1,000 and 1,900 give 400, 900 and 1,100 bp. The total is still 2,400 bp.
• All three together: cuts at 600, 1,000, 1,900 and 2,200 give 400, 900, 300 and 800 bp. Add them up and you get 2,400 bp again.

The tool takes a sequence, not a list of positions, so the positions above are what it would report for a plasmid with sites in those places.

On a gel, the 300 bp and 400 bp pieces are close, so a ladder with a band near 300 and 400 bp helps you tell them apart.

Where it is used

• Cloning. You open a vector with one or two enzymes that cut once in the place you want, then join your insert to it.
• Mapping an unknown plasmid. Several enzymes, each cutting a few times, give a pattern of sizes that points to where the sites lie.
• Checking a construct. After a ligation or mini-prep, the sizes on the gel show whether the insert went in.
• Teaching. Gel analysis is a standard school and university biology practical.
• Genetic testing. Restriction fragment length polymorphism (RFLP) tests were built on the same idea.

Common mistakes

1. Forgetting the plasmid is circular. Choose Circular. For a linear molecule the wrap-around piece does not exist, so the sizes are wrong.
2. Confusing a site with a cut. The site is the recognition sequence; the cut is a point inside it. When you type your own enzyme, put the ^ inside the site, at the point where the top strand is cut.
3. Expecting every site to be cut in the lab. Chemical tags on the DNA (methylation) can stop some enzymes from cutting. The screen does not know about them, so read your supplier’s notes before you trust a prediction.
4. Trusting one band for two fragments. Two pieces of nearly the same size run as a single band. Check the sum of all the pieces against the plasmid length.
5. Reading the virtual gel as a photograph. It places bands by a simple rule: the distance falls steadily with the logarithm of the size. A real gel also depends on the agarose percentage and the run, and an uncut, twisted plasmid does not run at its length.

Questions people ask

Why does my real gel show bands that the prediction does not? Common reasons are an incomplete digest (some DNA left uncut), extra cutting at near-matching sites under harsh conditions, or the wrong sequence on screen. The simulator assumes every site is cut fully and only at the exact site you gave.

Can I use an enzyme that is not in the list? Yes. Type it on its own line, with its site and the ^ position from your supplier’s data sheet.