From: Steve Kneizys <spk4@psu.edu>
  To  : RASMB@bbri.harvard.edu
  Date: Sun, 22 Mar 1998 21:47:57 -0500

se analysis of heteromeric systems

on Fri, 20 Mar 1998 18:44:29 -0500 (EST), Robert Fairman wrote:
>
>Does anyone have software for general analysis of heteromeric systems?  I
>particularly like the NONLIN program and was also wondering if any plans to
>include heteromeric models are in the works?
>

Funny you should ask...I am a 2nd-yr Biochemistry student at
the Penn State College of Medicine, and I have been without a lab 
since January.  While waiting for placement in a lab, I 
have upgraded software I had written previously to analyze
Calorimetry and NMR data to also handle Sedimentation Equilibria.
The program runs on PC's, Linux, and 68K Macs (my "Codewarrior"
is on backorder...PowerMacs soon.)

This email is going to be a little long because the software is
in development, and I would like any suggestions on design.  I
want it to be easy to use, and also useful.  This is a general
description so far:

Uses SIMPLEX method to minimize, Bevington's Uncertainty analysis.
Up to 15 data files of 350 points each.
Up to 6 monomers (A,B,C,etc.)
Up to 8 user-defined products

Of course, a system with 6 monomers reacting to product 8 products
would be impossible to sort out, but I didn't want to have too few
of these available.

Products are defined by the stoichiometry.  The software accepts
the numbers of each type of monomer involved in a complex.  For
example, in a system with A, B and C reacting that one was
expecting to see a 1:1:1 or 3:2:1 (or whatever) product, the
product is defined by these integers input from the keyboard.

The program will do a search for a 'pit' from initial guesses, 
but will also do a 'search-and-destroy', which is something 
that I have just started playing with (Basically a contour map
or schematic map.)

My problems are:

1) I am not in a lab, and have little access to test data.  
Does anybody out there have data from a well-behaved system I 
can do testing with?  (I am also willing to change grad 
programs...my department has only one open lab right now that
is not compatible interest-wise...I am just waiting for someone
to get a grant...I would relocate in a heartbeat!)

2) I do not know exactly how to define the interface, or which
parameters people would want to have as user-defined v. searched-
for.  Right now, in a binary system for example, the v_bar and 
MW are input, and then from the stoichiometry the MW and V_bar 
for the product are calculated (assumes ideality, beer's law.)  
Should the MW's be searched for?  Or should a previous velocity
run be used to fix these values?  

3) I am not sure how robust the software will be, although the 
Calorimetry/NMR code has been used extensively at Franklin and 
Marshall College...

Any test data & suggestions are appreciated!

thanks,

Steve...  
--
Steve Kneizys (spk4@psu.edu)
Graduate Student
Department of Biochemistry and Molecular Biology
Pennsylvania State University College of Medicine
The Milton S. Hershey Medical Center
P.O. Box 850
Hershey, PA 17033