Cot curves & RMS

GLEN SHEARER GLEN_SHEARER at BULL.CC.USM.EDU
Fri Jul 22 08:47:11 EST 1994


 
Perhaps someone out there who has suffered the slings and
arrows of Cot curves can lend some help.
 
We are doing DNA renaturation kinetics to help determine the
genome size & complexity of Histoplasma capsulatum.  We are using
an old copy of Dr. Pearson's program to determine the rate constant,
Cot 1/2 etc., which we are running on an old IBM PC.  We also
use the Macintosh graphics program Kaleidagraph to fit the data
to the same equation  %ss= f / (1 + k1 Cot).  This program is
faster and more user friendly.
 
Now, here is the question--- the Pearson program gives us an RMS
number which is usually fairly good (around 0.03).  The Kaleidagraph
program calculates a chi square value (usually aroung 0.02) and a
R value (usually around 0.99).  I am familiar with the R value
as a goodness of fit for curve fitting equations.  Does anyone
know how RMS compares to chi square or R?  I have looked through
every stat book I can find and asked several mathematicians.
Either they don't know or give me conflicting answers.
 
Many thanks for your help.
 
<glen_shearer at bull.cc.usm.edu>



More information about the Mycology mailing list