Find a motif

Identical to its reference in the validationA motif scan as seqkit locate makes it (Shen et al. 2016). doi:10.1371/journal.pone.0163962

In the app

Analysis panel, under 1D tasks: Find a motif

Command

chromosomix motif

Reads

Tracks and interval files, with or without a contact map

Also known as

sequence, pattern, sites, GATC, restriction

Find a motif finds every match of a motif written in IUPAC letters, such as GGTCANNNTGACC, in a loaded reference, as a set of sites and as a count of sites per bin. Use it for a binding-site consensus, a repeat or a telomere sequence; for where an enzyme cuts, use Restriction sites.

Running it

Load the reference as a FASTA or a .2bit (Reference sequences), open Find a motif on the 1D tasks card of the Analysis panel, choose the reference, type the motif and tick the sequences; no map is needed. A motif the search cannot read is refused, with the reason beside run.

The sites are a track on the Tracks card named after the search, such as GATC (both strands), and the sites per bin a second, the same name followed by density. With no map open there are no bins, so only the sites are made.

The same run from a shell:

chromosomix motif [<map>] --sequence <sequence.fa> --pattern <pattern> [--<parameter> <value> …] --out <dir>

Parameters

Case and spaces in the motif are ignored, so GGTCA NNN TGACC copied from a paper works as it is, and every overlapping match is counted. both strands is on, since a motif occurs on both strands of the molecule; a palindrome is still reported once. mismatches is for a consensus with variants; leave it at 0 for an exact string such as a telomere repeat. An N in the motif is always free.

Parameter	Takes	Default	What it does
reference--sequence <file>	a file: a reference sequence	—	The FASTA or .2bit to search — its own sequences, in its own order.
sequences--sequences <name,…>	sequence names, comma-separated	Every sequence of the file when left out.	—
motif--pattern <text>	text	None.	IUPAC — A C G T U and the ambiguity codes R Y S W K M B D H V N — with spaces and case ignored and every overlapping match counted.
both strands--bothStrands / --no-bothStrands	on or off	on	A palindrome is still reported once, not twice.
mismatches--mismatches <n>	a whole number, 0 to 4	0	Positions a site may differ from the motif by; N in the motif is always free.
most sites--limit <n>	a whole number, 1 to 500000	500000	A backstop against a pattern like N: up to 500,000, and the run says so when it bites.

What it writes

• intervals: a set of intervals: TSV, and BED (or BEDPE for pairs, or bed3d and PDB for a 3D model)
• profile: a value per bin: TSV, bedGraph and bigWig

On the command line every file goes into --out as <file stem>.motif.<output>.<ext>, and each TSV opens with # lines naming the build, the file and every parameter that produced it.

Reading the result

The sites are one interval per match, as long as the motif, named by the pattern and the strand; a palindrome searched on both strands is one site with no strand. With mismatches allowed, the score column holds how many positions each site differs by: filter on it to keep the exact matches.

The density is the number of sites in each bin, and a bin with no site is a zero, not a gap. Read the density at the scale of a chromosome, and the sites at a few kilobases.

What to watch for

• An N in the reference never matches a base the motif names, so an assembly gap is not a run of matches.
• One run reports at most 500,000 sites; a search that reaches the limit stops and says so. Narrow the motif, or tick fewer sequences.
• Each search is its own pair of tracks: another motif adds rows, and the same search run again updates its own.

Compared with seqkit

In the validation, the sites and strands are those of seqkit locate, except that seqkit reports a palindrome twice, once per strand.