How to measure or assign hydrophobicity score values to individual amino acids of a PDB structure?

Question

I want measure the hydrophobicity of each amino acid within a PDB structure file. Since I have the PDB file I want to consider the 3D information, rather than sequence-only measures such as GRAVY.

I am not interested in measuring the SASA (solvent accessible surface area) or RD (residue depth) - although correlated, it is not the information I require.

How can I comparatively measure the hydrophobicity of the amino acids within a PDB structure file? Is there an algorithm, library or program to do so?

Answer 1

No, this doesn't makes sense (I think).

The hydrophobicity of an AA is the hydrophobicity of the AA, in whatever context, just like it's molecular weight. It's weight is it's weight, in a structure or in isolation.

I know this seems physically unrealistic, but so is an abstract measure of 'hydrophobicity', it's just the best we can do based on partition coefficients.

Answer 2

Dan is correct but it doesn't answer the question. After much thought, I have finally figured out how to do it, only thanks to Dan's answer.

The problem is that a subsequence is one-dimensional and does not therefore include any of the additional information which is extractable from the three-dimensional structure (when available). Obviously, having the three-dimensional information is much more informative and descriptive for building classifiers, predicting interactions, interaction stability, or other purposes.

1. Find all intramolecular contacts within a radius of N. for the Amino Acid subsequence of interest (such as an Interface). I set N to, e.g. 4 Angstroms.

2. Calculate the descriptive statistics based on the set of hydrophobicity values for all intramolecular contact atoms or amino acids (depending on the granularity). e.g. Make an average of all of the hydrophobicity scores from the selected scale. (I use GRAVY, although there are many others).

To improve this further, the distance value of N can be set more intelligently than as a static value. For example, to the same as the cavity radius.