molecular surfaces

Friedrich Ackermann friedric at techfak.uni-bielefeld.de
Thu Apr 20 10:44:38 EST 1995


Dear Netters!
=============

I would like to invitate all those of you, that are working on
the fields of protein-protein-docking, molecular surfaces,
energy evaluation, electrostatic interactions in proteins etc.
to a little comparison of some results. Especially I would like
to compare the shape (geometry) of   m o l e c u l a r 
a c c e s s i b l e   s u r f a c e s   (MS) with each other. In 
addition, values of the electrostatic potential on the MS should 
be compared.
It is   n o t   intended to decide about the correctness of
underlying theoretical models, but rather to give all participants
a hint, how their results compare empirically with those of others.

My background:
==============
I am working on the problem of protein-protein-docking. The main
idea of our project is to integrate geometrical and chemical
surface features in the evalutation of complementarity in an
easy and fast manner. To this end I am developing a class library
(C++) to calculate and manipulate protein features and features
of the molecular surfaces. One of this features is the electrostatic 
potential on the surface. Now, the question arises: Am I computing
the right values? 

The comparison:
===============
Assume you can calculate the MS of a protein and possibly the 
electrostatic potential on it, i.e. you should be able to write down
a file similar to this:

x1 y1 z1 e1
x2 y2 z2 e2
x3 y3 z3 e3
x4 y4 z4 e4
...
xN yN zN eN

Hereby, xi, yi, zi, and ei are floats. xi, yi, and zi are the 
coordinates (in Aengstroem, given in the coordinate system of the 
PDB - entry) of a surface point and ei is the
electrostatic potential that you computed at this particular
surface point. Since we compare only relative values, it is not
important in which unit ei is given.
Your MS will be sampled in a 3D grid (isotropic lattice constants 
0.5, 0.8, and 1.2 Aengstroem will be used) and by counting hits and
missing voxels we will compare the shape (geometry) of the
different MSs. In a similar way we will compare by correlation on
the grid the values for the electrostatic potential on the MS.

You want to make one of the party?
==================================
Then please send me a note (friedric at techfak.uni-bielefeld.de)
(Name, institution, type of program you use, additional hints for
the comparison, references).
First all participants will be informed about the IDs of the others
and will receive a detailed description of the procedures used for 
comparison.
Then the PDB-IDs of some proteins I suggest to use for comparison 
and a description of possible file formats will be sent.
Afterwords you will calculate the MS and eventually the potential 
on it and you will send it to me. The comparisons will be performed, 
and the results will be posted to all participants as a set of upper
triangular matrices, containing cross correlations and similar values
of every MS with each other.
In addition the results will be compiled to a technical report, that 
will be made electronically available on our ftp server.

Schedule:
=========
5.5.1995	Deadline for note on intended participation.
10.5.1995	Posting of PDB-Ids and format descriptions.
		(Meanwhile eventually changes in procedure of 
		comparisons, depending on hints of participants.)
24.5.1995	Deadline for submission of MS
31.5.1995	Posting of results

I would be glad to here from you!
Friedrich Ackermann

THIS NOTE WAS POSTED BY ME ORIGINALLY TO THE FOLLOWING NEWSGROUPS:
bionet.biology.computational
bionet.molbio.proteins
bionet.software
bionet.announce
PLEASE FORWARD IT TO POSSIBLY INTERESTED PERSONS, THAT DON'T READ
THESE NEWSGROUPS!
PLEASE INFORM ME ABOUT OTHER NEWSGROUPS, INTO WHICH I SHOULD POST
THIS NOTE!






More information about the Bio-soft mailing list