Hydrophobicity plot

How to make a protein hydrophobicity plot?

Paste a protein sequence, FASTA records, or a PDB file, pick one of 24 amino acid scales, and set the window. The Hydrophobicity plot averages the scale values over every window of that many residues and plots each average at the window's center, so you can see which stretches of the chain are hydrophobic, hydrophilic, or, on the Welling scale, likely antigenic.

The Insulin example uses the Hopp-Woods scale with a 7-residue window:

>sp|P01308|INS_HUMAN Insulin OS=Homo sapiens OX=9606 GN=INS PE=1 SV=1
MALWMRLLPLLALLALWGPDPAAAFVNQHLCGSHLVEALYLVCGERGFFYTPKTRREAEDLQVGQVELGGGPGAGSLQPLALEGSLQKRGIVEQCCTSICSLYQLENYCN

Hopp-Woods rises with hydrophilicity, so the tallest peak marks the most water-exposed stretch. It reaches 2.01 at positions 56 and 57, in the charged TRREAED segment where the B chain joins the C-peptide, and the chart labels values above 0 as hydrophilic. On the default Eisenberg scale the same kind of profile is flipped: peaks are hydrophobic.

The first and last three residues have no full 7-residue window, so the plot runs from position 4 to 107. The values match ExPASy ProtScale for the same scale and window.

Input

Headers are kept whole as sequence names, and a plain name line above an unheaded sequence becomes its name. Spaces, numbers, and stop markers (*) are removed, and the page says when stop markers or alignment gaps were dropped. PDB files give one sequence per chain. Sequences shorter than the window are skipped with a note, and DNA pasted by mistake is flagged; translate it with DNA to Protein first.

Settings

Windows of 5 to 9 keep detail such as surface loops, and Hopp and Woods used 6 for antigenic sites. Windows near 19 match the length of a membrane-spanning helix.

Results

Export the chart as an image from the camera button. When the input has several sequences, choose one from the menu beside the chart.

The reference line sits at 0 for scales with positive and negative values. Seven scales never go below zero (Aboderin, Black-Mould, Chothia, Manavalan-Ponnuswamy, Miyazawa-Jernigan, Rao-Argos, and Rose), so zero would label every residue hydrophobic; for them the line sits at the mean of the 20 values, and the chart reads above it as more hydrophobic than average.

Rare residue codes follow ProtScale: B is the mean of D and N, Z the mean of E and Q, and X the mean of all 20 values. U, O, and J have no value on these scales, so windows that contain them average over their other residues, and the page lists them.

Available scales

All values match the scale pages of ExPASy ProtScale.

Choosing a scale

The 1.6 cutoff for transmembrane helices belongs to the Kyte-Doolittle scale with a 19-residue window; other scales have different ranges, so compare peaks within a plot rather than against that number. For a single number describing the whole protein, use the GRAVY calculator.

Which protein property tool should I use?

FAQ

Which hydrophobicity scale is best?

There is no single best scale. Eisenberg's consensus is a balanced default, Kyte-Doolittle is the usual choice for membrane helices, and Hopp-Woods or Parker suit surface and antigenic regions.

Is a positive value always hydrophobic?

No. On Hopp-Woods, Parker, Bull-Breese, and Guy, positive values are hydrophilic, and on Welling they mark antigenic residues. The chart legend states the direction for the selected scale.

Why does the plot start a few residues into the sequence?

Each point averages a full window centered on that residue, so the first and last (window - 1) / 2 residues have no value. This matches ExPASy ProtScale.

Can a hydrophobicity plot predict epitopes?

It can point to hydrophilic, surface-exposed stretches that antibodies often bind, which is what Hopp-Woods and Parker were designed for. It does not account for structure or sequence conservation, so treat peaks as candidates.