Q.843 - How can I calculate the volume of a protein ?

Category: S.843 - general

A simple question triggered a lot of responses: "Is there an easy way to find the volume of a protein molecule ?"

Eleanor Dodson replied: "A rough but fairly accurate method is to count the number of atoms - CCP4 rwcontents reads a PDB file; use some approximation. Add on the number of hydrogens ( ~ = number of non- hydrogens) then say each atom has a volume of ~ 10A**3. Agrees with measured densities pretty well."

C.S. Raman suggested: "You might want to try the program "VOLUME" distributed by Fred Richards at Yale U. See J.Mol. Biol. (1974) 82: 1-14 & Met. Enzymol. 115:440-464 (1985) for details. In its present implementation, "VOLUME" caclulates the volume of a polyhedron surrounding each specified atom(s) of a protein where the polyhedral faces are constructed based on the Voronoi method. You can obtain a copy of the license agreement from Art Perlo also at Yale."

Ruslan retorted: "I have read the two answers to the question and still seems to me that the EASYEST way is to do this with the Matthews Number provided, of course, that one has crystals of the protein."

Janet Smith wrote the following essay: "The volume of a protein molecule can be approximated very simply and reliably from the molecular weight of the protein and an average protein partial specific volume. (Partial specific volume = volume / molecular weight.) The average of experimentally determined partial specific volumes for soluble, globular proteins is ~0.73 cm**3/g. This value varies from protein to protein, but the range is rather narrow.

The simple calculation starts from

	0.73 cm**3/g  x  10**24 A**3/cm**3  x  molecular weight g/mole
	--------------------------------------------------------------
			 6.02 x 10**23 molecules/mole

and results in a protein volume of ~(1.21 x MW) A**3/molecule. Volumes estimated in this way are quite accurate for our usual purposes. For example, when compared to the volume of a crystallographic asymmetric unit, it is very simple to estimate the number of protein molecules or subunits in the asymmetric unit or to compute the solvent content of crystals.

For a recent, brief tabulation of partial specific volumes, see Harpaz, Gerstein and Chothia (1994) Structure 2, 641-649 (issue of July 15). Their Table 2 lists partial specific volumes between 0.70 and 0.75 cm**3/g for 13 soluble proteins."

Pete Dunten played by the rules: "Hey, this is the o-bulletin board, and no one has proposed a solution making use of o yet!!!! SO, IF WE PLAY BY THE RULES, we would read the pdb into MAMA, make a mask, and take note of the volume enclosed by the mask.

OR, making better use of our fancy graphics facilities and o itself, we would display the molecule, and then use distance_define to take down the dimensions across the molecule, and proceed from there with a simple back-of-the-envelope calculation."

Gerard noted briefly: "VOIDOO quickly calculates volumes which in my experience are good enough for government work ;-)"