Surface accessibility
Measures how exposed each residue is to solvent, as an absolute area and as a fraction of that residue type's maximum. Use to tell surface from buried before choosing mutation sites, to find an exposed patch, or to see which residues a partner would cover.
Surface accessibility
What it does
Takes one structure and reports, for every residue, how much of it is reachable by solvent. Two numbers per residue: the absolute area in square angstroms, and the relative figure — that area as a fraction of what the same residue type would have if it were fully exposed.
The relative number is the one to read. 50 Ų is buried for a tryptophan and almost fully exposed for a glycine, so the raw area alone tells you little.
How the area is calculated
The engine is FreeSASA, using the Lee–Richards definition: roll a probe the size of a water molecule (1.4 Å) over the surface and record where its centre can reach. Anything the probe cannot touch is buried.
The calculation is done on the file as given. If two chains are present, a residue at the interface between them counts as buried, because the partner is there. Running one chain on its own is how you see what that residue would look like unbound — and the difference between the two is the interface.
Reading the output
Relative exposure above about 25% is the usual line between surface and buried, and is what the summary counts. The table lists the most exposed residues first, since "what is on the surface" is normally the question.
Typical uses:
- Choosing a mutation site — a buried residue is usually structural, and changing it risks the fold rather than the function
- Finding an exposed patch — a run of exposed hydrophobics often marks where something else binds
- Measuring an interface — run the complex and each partner separately, and compare
Relative values can exceed 100% for terminal residues, which have more exposed area than the reference value allows for. That is expected, not an error.
Important limitations
It is a static calculation on one conformation. A side chain that moves will have a different answer in a different structure of the same protein.
Missing atoms read as absent, not as buried. If a side chain was not resolved, its area is simply not counted.
Waters and ligands in the file are part of the surface unless removed, so a residue under a bound ligand will read as buried.