BioDeviceHub

Cas9 Target Finder and PAM Finder

Find Cas9 target sites in a sequence on both strands: the PAM (NGG, NG or NNGRRT), the adjacent protospacer, its GC content, the cut position and basic flags.

Formula

SpCas9 PAM: NGG directly 3′ of a 20 nt protospacer\text{SpCas9 PAM: NGG directly 3}'\text{ of a 20 nt protospacer}
SaCas9 PAM: NNGRRT after a 21 nt protospacer\text{SaCas9 PAM: NNGRRT after a 21 nt protospacer}
cut: 3 bp upstream of the PAM, blunt\text{cut: 3 bp upstream of the PAM, blunt}
PAM\mathrm{PAM}
protospacer adjacent motif, the short sequence the nuclease needs next to its target
protospacer\text{protospacer}
the target sequence matching the guide's spacer

How it works

Cas9 cuts only where its target is followed by a PAM. For the commonly used SpCas9 this is NGG, so a target site is any 20 bases lying immediately 5′ of an NGG. The tool finds every such site in the sequence you paste, on both strands, and reports where the nuclease would cut, three base pairs upstream of the PAM.

The engineered SpCas9-NG recognises a shorter NG PAM and so has far more sites, though it tends to act less efficiently at some of them. SaCas9 has a longer NNGRRT PAM and a slightly longer spacer. Each site is flagged for TTTT (a Pol III terminator) and for GC outside 40 to 80%, and for copies of the same protospacer within your sequence. Finding a site is the start of guide design, not the end.

Worked example

An illustrative 91 bp sequence, SpCas9 with a 20 nt spacer.

  1. Every NGG on the top strand with 20 bases before it is a site, and every CCN on the top strand is an NGG site on the bottom strand.
  2. The first site on the top strand is at 18 to 37, with PAM CGG; Cas9 cuts 3 bp upstream of the PAM, after base 34.

Ten target sites across the two strands, none flagged.

These are the values the calculator opens with, so you can check its output against this example.

Assumptions

  • Sequence matching only, to the PAM and the protospacer length you choose.
  • Sites are searched only in the sequence you paste, not across a genome.
  • No cutting efficiency or specificity is predicted.

Common mistakes

  • Treating a site with a matching PAM as a good guide. Efficiency varies enormously between sites.
  • Skipping a genome-wide off-target search, which is essential for guides used in cells or animals.
  • Forgetting that a target whose PAM is destroyed by the edit will not be re-cut, and one whose PAM remains will.