Restriction Site Finder

How to Use the Restriction Site Finder

1. 1. Paste the sequence. Plain letters or FASTA (the first record is used), up to 100,000 bases. Choose Circular for a plasmid and Linear for a fragment or PCR product.
2. 2. Choose the enzymes. The popular set covers the cloning enzymes most labs stock; All NEB enzymes tests every enzyme on the list. You can also type your own recognition site.
3. 3. Scan the table. For each enzyme you see how many times it cuts, the cut positions, the recognition site and the ends it leaves. Filter to single cutters, all cutters or non-cutters.
4. 4. Digest. Tick one or more enzymes to get the fragment sizes, a map, a list of every site with its coordinates, and the sequence with the sites highlighted.

How Restriction Enzymes Recognise and Cut

A restriction enzyme binds a short DNA sequence and cuts both strands at fixed positions. Most cloning enzymes are Type II: their recognition sites are palindromic — the sequence reads the same 5′ to 3′ on both strands — so one enzyme cuts both strands symmetrically. EcoRI recognises GAATTC and cuts G^AATTC, leaving a 5′ overhang of AATT on each fragment. Enzymes such as KpnI (GGTAC^C) and PstI (CTGCA^G) leave 3′ overhangs, and SmaI (CCC^GGG) leaves blunt ends. Fragments with compatible overhangs can be joined by DNA ligase, which is the basis of restriction cloning.

Type IIS enzymes cut outside their recognition site. BsaI binds GGTCTC and cuts 1 base beyond it on the top strand and 5 on the bottom, as GGTCTC(1/5), leaving a 4-base overhang whose sequence you choose. That is the property Golden Gate assembly exploits. Many enzymes recognise degenerate sites written with IUPAC codes — HinfI cuts G^ANTC, where N is any base, and BsaWI cuts W^CCGGW — and some, like BglI (GCCNNNN^NGGC), have a gap of unspecified bases in the middle. For a site that is not palindromic, the enzyme also cuts where the reverse complement appears, so the tool searches both strands and lists such sites with strand −.

Reading Cut Positions

Cut positions are given as cut after base N on the top strand: in the pUC19 sequence below, EcoRI cuts after base 396 and HindIII after base 447. (Biopython's Restriction reports the first base after the cut, one higher.) Each site's start and end are listed too. Fragment sizes are the distances between those top-strand cuts, so the overhang is counted in whichever fragment lies to its 3′ side. On a circular sequence the last fragment wraps around the origin, and n cuts give n fragments; one cut simply linearises the plasmid. On a linear sequence a Type IIS cut that would fall beyond either end is ignored.

Worked Example: Mapping pUC19

Try an example loads the 2,686 bp cloning vector pUC19 (GenBank L09137.2) as a circular sequence. Of the 66 popular enzymes, 36 cut it and 19 cut exactly once; across all 229 NEB enzymes on the list, 152 cut and 46 are single cutters. Ten of the unique cutters sit together in the multiple cloning site between bases 396 and 447:

Digesting with EcoRI and HindIII together releases a 51 bp fragment and leaves a 2,635 bp vector backbone (51 + 2,635 = 2,686). PstI and SphI, only six bases apart, give 6 bp and 2,680 bp. Because the cut sites are unique, any pair lets you drop an insert into the vector in a defined orientation — the idea behind directional cloning. These positions agree with Biopython's Restriction.search for every enzyme tested, in linear and circular mode.

What the Tool Does Not Know

• Methylation. Dam (GATC), Dcm (CCWGG) and CpG methylation block or permit some enzymes: DpnI cuts only methylated GATC, while MboI is blocked by it. A site listed here may not cut in DNA from a methylating E. coli strain.
• Star activity and buffers. Enzymes can cut at degenerate sites under non-optimal conditions, and compatibility with a buffer is a separate question.
• Which enzymes you can buy. The list is the NEB-supplied enzymes in the REBASE database (Roberts et al., rebase.neb.com). Enzymes that cut on both sides of their site (BaeI, BcgI, BsaXI, CspCI), homing endonucleases, EcoP15I and MspJI are left out. Check the supplier's catalogue before ordering.

Frequently Asked Questions

How do I find restriction sites in a DNA sequence?

Search the sequence for each enzyme's recognition sequence, allowing for ambiguity codes, on both strands. Paste the sequence here and the tool does that for dozens of enzymes at once, listing each site with its coordinates and the position where the enzyme cuts.

What does the caret (^) mean in a restriction site?

The caret marks where the enzyme cuts the top strand. In G^AATTC, EcoRI cuts between the G and the first A; the bottom strand is cut symmetrically, which leaves the 5′ AATT overhang. Type IIS enzymes use (n/m) instead, giving the cut positions on the top and bottom strands as bases beyond the end of the site.

What is the difference between sticky and blunt ends?

Staggered cuts leave single-stranded overhangs, called sticky ends, that base-pair with any fragment carrying a complementary overhang. A cut straight across both strands gives blunt ends, which ligate to any other blunt end but more slowly and without orientation control. Enzymes with the same overhang, such as BamHI and BglII (both GATC), produce compatible ends.

How do I choose enzymes for cloning?

Pick enzymes that cut the vector once in the cloning site and do not cut your insert. Use two different enzymes with different overhangs for directional cloning, so the insert can enter in only one orientation. List the unique cutters in the vector, then check each against the insert sequence.

What is a Type IIS restriction enzyme?

A Type IIS enzyme cuts at a fixed distance outside its asymmetric recognition site, so the overhang it leaves is not part of the site and can be any sequence. BsaI, BsmBI, BbsI and SapI are used for Golden Gate assembly because the ligated product loses the recognition site and cannot be re-cut.

Should I treat my plasmid as circular?

Yes, for an intact plasmid. A circular molecule has as many fragments as cuts, and a recognition site can span the point where the sequence is numbered from. Choose linear for a PCR product or a fragment with free ends.