Sequence search
TL;DR: The linear genome view's Sequence search menu item searches the reference sequence itself and adds the hits as a new track, whether or not a sequence track is displayed. Three modes:
- Sequence pattern - find every occurrence of a single DNA motif or regex across the reference.
- CRISPR guide RNAs - discover candidate guides and their PAM sites directly against the reference.
- Motif list - search many named motifs at once, pasted as one per line. This is the mode for restriction enzymes; see below.
Sequence pattern
Enter a sequence to find every occurrence in the reference. The field is a
regex, so GAATTC matches exactly and TATA[AT]A[AT] matches either variant.
Both strands are searched, case-insensitively; either can be turned off.
Most restriction sites are palindromic — the site reads the same on both
strands, which is how one enzyme cuts both — so such a hit is reported once,
unstranded, rather than twice at one position. This needs bare ACGT: regex
syntax has no reverse complement, and IUPAC codes are not regex (N matches a
literal N). Use Motif list for
IUPAC sites.
CRISPR guide RNAs
Choose an enzyme preset (SpCas9 NGG, SaCas9 NNGRRT, Cas12a TTTV, ...) or
set the PAM, guide length, and PAM location by hand, and search either or both
strands. Each guide is drawn with a dedicated guide-RNA glyph: the protospacer
box, the PAM overpainted, and the predicted cut — one tick for a blunt cutter
like SpCas9, a staggered pair for Cas12a.
Clicking a guide reports its sequence, PAM, cut position, GC%, and two flags.
hasPolyT marks the TTTT run that terminates transcription from the pol III
promoters guides are usually expressed from. softMasked means the reference
lower-cased this protospacer, which marks it as repetitive. Both describe the
protospacer's own sequence; scoring a guide's specificity takes a genome-wide
search, and this mode runs a scan of the region in view. A protospacer that
reaches into an assembly gap is skipped.
To order oligos from a search, use Save track data on the track's menu. The GFF3 export carries the guide sequence, PAM, cut positions, and flags as attributes.
Restriction enzymes and other named motifs
The Motif list mode takes a pasted list in the notation REBASE already uses, one motif per line, with an optional name and an optional cut marker:
# name site
EcoRI G^AATTC
BamHI G^GATCC
PstI CTGCA^G
BsaI GGTCTC(1/5)There are two cut notations because there are two kinds of enzyme. A ^ marks a
cut inside the recognition site. The (n/m) form is for the type IIS enzymes
that cut downstream of their site — BsaI, BsmBI, BbsI, SapI, AarI, the Golden
Gate workhorses — where n and m count from the site's 3' end to the top- and
bottom-strand cuts (negative numbers count back into the site).
The panel is prefilled with common cloning enzymes; wipe it and paste your own
set from REBASE or anywhere else. Sites may use
IUPAC ambiguity codes (e.g.
GGTNACC), blank lines and # comments are ignored, and a bare site with no
name names itself.
The cut notation carries the cuts themselves. (n/m) pins both strands' cuts
outright; a ^ pins the top-strand cut, and for a palindromic site the
bottom-strand cut mirrors it. So each hit reports both cut positions and whether
the enzyme leaves a 5' overhang, a 3' overhang, or a blunt end. The strand
checkboxes only appear when the list contains a motif that is actually stranded.
Because the list is just text, the same mode works for primers, adapters, polylinker sites, or any other named motif set.
With more than one motif in the list, Launch as one track puts every motif's matches in one track, distinguishable only by name, and Launch one track per motif gives each its own lane.
Ship a search in config.json
A search launched from the menu lives in your session. To put one in front of
everyone who opens the instance, write the same adapter into config.json as an
ordinary FeatureTrack — it names no file, since the assembly supplies the
sequence. The cookbook has a whole track config
for each of the three modes.