What are the units for pep_patch_electrostatic patch areas?

[DavidBJaffe]

Thank you for creating this wonderful program!

I've been running pep_patch_electrostatic on Fv parts of antibodies using -o patches.csv. The area column in patches.csv typically maxes out at 4 or 5. But what are the units?

From reading the paper, I thought it was in Å^2. And in the code I see

        '-s','--size_cutoff',
        type=float,
        default=0.,
        help='Restrict output to patches with an area of over s A^2.

which makes me think the area is indeed in Å^2. However, the reported area values seem about 100 times too small for that to make sense.

Elsewhere in the code I see

    def as_plydata(self, text=True, units_per_angstrom=0.1):
        """Convert to a plyfile.PlyData object, while scaling coordinates

        the default scaling units_per_angstrom=0.1 matches the PyMol CGO
        scaling factor of 1:100 and the nm -> A conversion.

which might explain the factor of 100.

So are the units for the reported area values actually (10^-7m)^2, rather than Å^2?

[fwaibl]

Hello,

thanks for bringing this up! Yes, the units should indeed be nm^2. We had inconsistent units in the original version of our program, and changed it recently to use nanometers everywhere, but obviously we forgot to update the help message. I will change this message as soon as possible.
The PLY file has different units, to display correctly in Pymol. I don't remember what the final unit was in the PLY, but it might be the (10^-7m)^2 that you suggested. However, none of the numbers in the output have this unit, they should all be in nm.

[DavidBJaffe]

My suggestion is that the area values in the output file are 100 times too small. If one looks at the PLY files in Pymol one sees marked regions that are clearly much larger than the outputted areas indicate. Also in the literature, area values for charge patches are much larger than those produced by pep_patch_electrostatic. For example, in 2024. Armstrong GB et al. PMID 38867721, the values that are shown for individual residues are indeed ~100 times larger than the values emitted by PEP-Patch, and taking account of that, generally agree within a factor of ~2.

For this to happen, the units for area would have to be (10^-7m)^2, and notnm. To actually prove this, I suppose one would have to carefully read through the code, which I haven't done.

[fwaibl]

But the conversion is 1 nm^2 = 100 A^2. Wouldn't that explain the factor you are seeing?

Best,
Franz

[DavidBJaffe]

Sorry, I was being completely dense. You're right!