Initiation-window fold

The mRNA secondary structure around a start codon, drawn as arcs on the sequence axis, so an occluded ribosome binding site is something you can see rather than infer from one energy.

What it doesCopy Anchor Link

A ribosome does not read a straight line. Before initiation, the messenger has already folded on itself, and whether the 30S subunit can reach the Shine-Dalgarno sequence and the start codon depends on what that folding covers. Two constructs with the same Shine-Dalgarno sequence, the same spacing and the same start codon can express very differently for no reason other than a hairpin that one of them forms across the site.

The initiation-region free energy on the strength readout is that effect, reduced to one number. It tells you the structure is costly; it never tells you what the structure is, so it cannot tell you which bases to change. This is the same fold drawn out.

Each arc joins two paired positions. The window is the 35 nucleotides either side of the start codon — the footprint that matters for initiation, and the same window the reported energy is computed over, so the picture and the number describe the same bases rather than two regions that happen to overlap. The start codon is marked on the axis. A stem whose arc springs over that mark, or over the Shine-Dalgarno region upstream of it, is structure the ribosome has to compete with.

The layout is linear and coordinate-preserving on purpose: nucleotide i stays at position i, so the drawing lines up with the sequence axis and with every other track in the editor. Circular and force-directed layouts are prettier in isolation and cannot be read against a coordinate.

When you'd use thisCopy Anchor Link

• When a construct expresses poorly and the Shine-Dalgarno sequence looks right. A strong site that is base-paired shut is a different problem from a weak site, and only one of them is fixed by changing the Shine-Dalgarno sequence.
• When you are choosing where to make a silent change. The arcs name the partner of every paired base, so you can see which codon at the 5′ end of the coding sequence is holding the stem closed.
• When comparing two 5′UTR designs. A lower initiation-region energy means more structure, which is worse for expression — the sign is easy to read backwards from a single number and hard to read backwards from a picture.
• After a reverse-designed 5′UTR. The design targets a strength; folding the result shows whether the sequence it chose introduced structure of its own.

What this computesCopy Anchor Link

• The minimum-free-energy secondary structure of the initiation window — a nearest-neighbour thermodynamic fold using the Turner 2004 parameter set (Mathews et al., PNAS 101
  –7292), reported as the pairing partner of every position in the window.
• ΔG of that fold, in kcal/mol — the folding free energy of the window itself. More negative means more stable structure over the initiation region, which lowers initiation rather than raising it. The relationship between the stability of structure at a ribosome binding site and translational efficiency is the one de Smit and van Duin measured (PNAS 87
  –7672, 1990).
• The position of the start codon within the window, marked on the axis, so occlusion is read against the site it would occlude.
• The window's span on the sequence you have open — the same coordinates the map highlight draws, taken through the strand-aware mapping rather than re-derived, so a coding sequence on the bottom strand reports where it actually is.

The Shine-Dalgarno hybridisation energy, the spacing and the relative strength are separate readouts on the strength panel. They come from the same model; this view is the structural term of it.

What it does not doCopy Anchor Link

It folds a bounded window, not the transcript. Long-range pairing between the initiation region and a distant part of the message is real and is outside what a 70-nucleotide window can see. It is a minimum-free-energy structure, so it reports one structure and not an ensemble: a sequence that populates two comparable conformations is drawn as whichever is lower, with no indication that the other exists. It is a prediction at equilibrium, and translation is not at equilibrium — a ribosome that initiates unwinds what it reads.