Restriction Enzyme Site Finder
A restriction enzyme site finder scans a DNA sequence you paste and reports every enzyme that cuts it, where each cut falls, and the fragment sizes the digest would produce. This one matches your sequence against the complete REBASE set — 5,649 enzymes — entirely in your browser, with no upload and no account. For each cutter it lists the recognition site, the number of cuts, the break positions, and the resulting fragments for a linear or a circular molecule, so you can see at a glance which enzymes linearize a plasmid once, which cut it into a clean two- or three-band pattern, and which do not cut at all.
Every cloning strategy begins with a question: which restriction enzymes cut this molecule, and what do the fragments look like? Answering it by hand means scanning a sequence for dozens of recognition motifs and then doing the fragment arithmetic for each one, which is slow and easy to get wrong — especially across the join of a circular plasmid or for the offset cuts of a type-IIS enzyme. This tool is part of the molecular cloning calculator set and does the whole scan in one pass, listing every cutter with its recognition site, cut count, cut positions and digest fragment sizes.
The engine reads recognition data derived from REBASE, the reference database of restriction enzymes maintained by Roberts and colleagues, so the enzyme set is the full published collection rather than a short list of common clones. That coverage matters when your construct carries an unusual site: a tool limited to a few hundred commercial enzymes will simply never report a cutter it does not know, and you will wrongly conclude the site is unique. Because the scan runs locally, you can check a proprietary or unpublished sequence without sending it to a server.
Reading the output well requires understanding what a cut position and a fragment size actually mean. A recognition sequence is written with a caret where the enzyme cuts the top strand — EcoRI is G^AATTC, so the break falls one base into the site — and type-IIS enzymes such as BsaI are written GGTCTC(1/5), cutting a fixed distance outside the site. Once you know where the breaks fall, the fragments follow simple arithmetic: a linear molecule cut n times gives n+1 fragments, while a circular molecule cut n times gives n fragments, and a single cut on a circle merely linearizes it. If you need the mass of those fragments rather than their lengths, the plasmid molecular weight calculator converts size to molecular weight.
The usual next step is to pick an enzyme and build a ligation around it. Choosing between two candidates that both cut your vector once comes down to the overhang they leave and whether the insert carries a compatible end; the ligation ratio calculator then turns your chosen insert and vector sizes into the masses to pipette. If you are designing the primers that add those sites in the first place, the DNA and RNA sequence tools cover melting temperature, GC content and molecular weight for the oligonucleotides and the final construct.
Restriction Enzyme Site Finder
How to use the restriction enzyme site finder
Paste your sequence into the input box. Only the strict bases A, C, G, T and U are accepted; spaces, line breaks and position numbers are ignored, and U is read as T so an RNA primer works. Degenerate IUPAC codes such as N are not accepted in the query, because an ambiguous base in your sequence would make the fragment sizes ambiguous too.
Choose the topology. Leave it on Circular for a plasmid or any covalently closed molecule so the scan searches across the origin; switch to Linear for a PCR product, a restriction fragment, or a linearized vector, where the ends are real ends and a cut near either terminus yields a short end fragment.
Press Find sites. The enzyme database loads on the first click, then the full REBASE set is scanned against your sequence. Nothing is uploaded — the matching happens in your browser.
Read the hit table. Each row is one enzyme that cuts, showing its recognition site, the number of cuts, and the fragment sizes in base pairs, sorted so the pattern is easy to compare against a gel. Enzymes that do not cut are omitted.
Expand a row for detail: isoschizomers, commercial availability where the source data records it, and the exact cut positions. Use this to confirm whether a site is truly unique or shared with an isoschizomer.
Copy the summary when you want to paste the result into a notebook or a protocol.
Restriction Enzyme Site Finder formula
Variables and units
| Symbol | Meaning | Unit |
|---|---|---|
| pattern | Recognition sequence, IUPAC-expanded for degenerate positions | bases |
| cutTop | Top-strand cut offset from the 5′ end of the site | bases |
| cutBottom | Bottom-strand cut offset from the 5′ end of the site | bases |
| cut position | Absolute base index of a double-strand break in your sequence | bp |
| fragments | Lengths of DNA between consecutive breaks, for the chosen topology | bp |
| topology | Circular (searched across the origin) or linear (real ends) | — |
How restriction enzyme site finder is calculated
- Normalize the input: strip whitespace and digits, uppercase, and read U as T.
- Validate strictly against A, C, G, T, U and reject the sequence if it exceeds the 100,000 bp scan cap.
- For each enzyme, expand its recognition pattern into per-position base sets (IUPAC-aware).
- Match the pattern on both strands: the forward strand directly and the reverse strand via the reverse complement of the pattern, searching across the origin when the topology is circular.
- Deduplicate palindromic self-matches so a site found on both strands at one locus counts once, keyed by its break position.
- Compute each double-strand break as site start + cutTop, wrapped into the sequence length.
- Derive fragments from the sorted break positions using the linear (n+1) or circular (n) rule, joining across the origin for circular molecules.
- Sort the fragments ascending and the enzymes alphabetically, then render the table.
Worked laboratory examples
Example 1 — EcoRI on a 60 bp plasmid (circular)
Given: 5′-TTAAGAATTCGGCCGGATCC…-3′ (60 bp) · topology: circular · EcoRI site GAATTC at index 4
- EcoRI is written G^AATTC, so cutTop = 1.
- The top-strand break is at site start + cutTop = 4 + 1 = 5.
- A single cut on a circular molecule does not produce two pieces — it linearizes the circle.
Result: EcoRI: 1 site, 1 fragment of 60 bp (the circle opened once).
Example 2 — the same sequence, linear
Given: 5′-TTAAGAATTCGGCCGGATCC…-3′ (60 bp) · topology: linear · EcoRI break at position 5
- A linear molecule cut once gives n + 1 = 2 fragments.
- The break at position 5 splits the 60 bp into 5 bp and 60 − 5 = 55 bp.
Result: EcoRI: 2 fragments, 5 bp + 55 bp.
Example 3 — BsaI type-IIS offset cut (linear)
Given: 5′-AAAAAAAAAA GGTCTC AAAAAAAAAAAAAA-3′ (30 bp, linear) · BsaI written GGTCTC(1/5) · site GGTCTC starts at index 10
- The (1/5) form sets cutTop = pattern length + 1 = 6 + 1 = 7.
- The break is therefore 7 bp past the site start: 10 + 7 = 17, outside the recognition sequence.
- A single cut on this 30 bp linear molecule gives fragments of 17 bp and 30 − 17 = 13 bp.
Result: BsaI: 1 site, break at position 17, fragments 17 bp + 13 bp.
Want the reasoning behind the numbers? Read How to Calculate Plasmid Molecular Weight and How to Calculate a Ligation Ratio (Insert:Vector).
How to interpret the result
A result where many enzymes cut once or twice is typical for a small plasmid; the useful signal is the enzymes that cut exactly once, because those linearize the vector cleanly for a directional clone, and the enzymes that cut zero times, which confirms a site is absent and therefore available to engineer in.
Circular versus linear changes the fragment count even for identical cuts. One EcoRI site gives a single 60 bp band on a circle but two bands, 5 and 55, on a linear molecule. If you are checking a plasmid prep, keep the topology on circular so a site spanning the origin is not missed; if you are checking a PCR product or a digested insert, use linear so the short end fragments are reported correctly.
A single-cut linearization is what you want when opening a vector for ligation: it produces one full-length linear molecule with two compatible ends. Two cuts excise the fragment between them, which is how you drop out a stuffer or release an insert. Read the fragment sizes against your expected construct — an unexpected extra band usually means a second site you did not plan for, or a partial digest.
Once you have chosen an enzyme, carry its fragment sizes forward. The linearized vector length and the insert length feed the ligation ratio calculator to set the molar ratio, and the vector length feeds the plasmid molecular weight calculator when you need mass rather than base pairs. Together these turn a site scan into a complete cloning plan.
Assumptions
- The query is a strict ACGTU sequence; degenerate bases in your input are rejected rather than guessed.
- The enzyme set is the complete REBASE collection used to build the database, applied to every sequence identically.
- Cutting is ideal: every recognition site is assumed to cut fully, with no methylation blocking, star activity, or buffer and temperature effects.
- The default topology is circular, matching the common case of a plasmid; switch it explicitly for linear molecules.
Limitations
- Only single-enzyme digests are reported; planning a combined double or multi-enzyme digest is a separate task.
- Methylation sensitivity, star activity, and buffer or temperature compatibility are not modelled, so a listed cut is a prediction of sequence recognition, not a guarantee under your conditions.
- Sequences longer than 100,000 bp are rejected rather than truncated; use a dedicated genome-scale tool for whole chromosomes.
- Overhang composition is not reported — only fragment sizes — so compatible-end matching must be checked separately.
Common mistakes
- Forgetting to set circular versus linear, which changes the fragment count and can hide a site that spans the origin of a plasmid.
- Assuming every listed enzyme will cut in your buffer; recognition is necessary but not sufficient, and buffer, methylation and temperature decide the real outcome.
- Ignoring methylation-sensitive sites: a recognition sequence inside a methylated host context may not cut even though the finder reports it.
- Reading a type-IIS offset such as GGTCTC(1/5) as a cut inside the recognition site; the break falls outside it, which is exactly what makes Golden Gate overhangs programmable.
- Pasting a sequence with N or other degenerate bases and expecting a result; ambiguous positions make fragment sizes ambiguous, so the tool requires strict ACGTU.
FAQ
What is a restriction enzyme site finder?
How are digest fragment sizes calculated?
Circular versus linear — why does it matter?
Which enzymes are included?
Is my sequence uploaded anywhere?
Why can’t I paste N or degenerate bases?
Scientific Sources
- REBASE — Roberts et al. (2023) Nucleic Acids Res. 51:D629–D630 (v610 enzyme data) · Tier 1 source
- PubMed 36318248 — REBASE: a database for DNA restriction and modification · Tier 1 source
- Wikipedia — Restriction enzyme · Tier 3 source
External links open in a new tab and carry rel="nofollow". We cite primary references (NIST, NIH/NCBI, peer-reviewed papers) and established manufacturer technical documentation (NEB, Thermo Fisher, Promega, IDT). Calculated results are estimates; your protocol and manufacturer guidance take precedence.
Last reviewed: 2026-09 · Method version: 1.0