Hydropathy profile
A residue-property scale averaged in a sliding window along the chain, so a buried core, a membrane-spanning stretch or an exposed loop is read as a position rather than as an average.
What it does
A protein's mean hydropathy is one number for the whole chain, and almost nothing follows from it. What follows from the profile is where the chain is hydrophobic: a run of twenty residues above the line in the middle of a soluble protein is a candidate transmembrane helix; the same mean spread evenly is a protein with an ordinary core.
The plot is a scale value per window centre, along the residue axis. Every published residue-property scale in the set can be drawn on it — hydrophobicity in its many parameterisations, but also bulkiness, polarity, flexibility, refractivity, accessible surface area, antigenicity, and the transfer energies used for membrane prediction — and several can be overlaid to see whether they agree about where a feature is. When they disagree, the disagreement is the finding: scales derived from partitioning experiments and scales derived from structural statistics do not measure the same thing, and a peak present in one and absent in another is telling you which.
The window length is yours. A short window resolves single residues and is dominated by them; a long window is what makes a helix-length feature stand out from the noise, at the cost of blurring its edges. Nothing here picks one for you, because the right window is a function of what you are looking for.
When you'd use this
- To find candidate membrane-spanning segments before committing to a topology prediction: a sustained hydrophobic run of roughly helix length is the signal every such method starts from.
- To locate a signal peptide or an anchor at the termini, where a short strongly hydrophobic stretch sits against a very different remainder.
- To choose where to truncate a construct. Cutting inside a hydrophobic core is how a soluble domain becomes an inclusion body.
- To compare two scales deliberately. Overlaying is how you find out whether a feature is a property of the protein or of the parameterisation.
- To read flexibility rather than hydropathy, which is a different question about the same axis: where the chain is predicted to move.
What this computes
- A windowed average of the selected residue-property scale, at each window centre along the chain, with optional edge weighting so residues near the window's edge count for less than those at its centre.
- Optional normalisation to the 0–1 range of the scale in force, which makes two scales with different units comparable in shape while destroying their absolute values — so the zero line is not drawn when it would be meaningless.
- A zero reference line, drawn only when the scale in force actually takes both signs. A scale that is positive everywhere has no zero to cross, and a line under the whole curve would invite reading one.
- The extremes and where they fall — the maximum and minimum window and the residue positions they centre on, stated in the reading beneath, because a curve does not give you a number to quote.
- The overlaid scales' peaks and how far they are apart, when more than one is drawn.
What it does not do
A profile is a property of the sequence, not a structure prediction. A hydrophobic run is consistent with a transmembrane helix, with a buried core strand, and with an aggregation-prone stretch, and this plot does not distinguish them — that is what the topology and amphipathicity readings are for. The window average also has no notion of where a domain begins, so a feature that spans a boundary is drawn straight through it.
Primer binding track
Where the selected pair binds on the template and what it produces, drawn on the template's own coordinates.
Charge vs pH
The titration curve of the protein: net charge across the pH range, with the isoelectric point where it crosses zero and the charge at the pH you actually work at.