
INFORMATION FOR THE USE OF THE RBINDING AND RBG PROGRAMS
FOR ANALYSIS OF RECEPTOR-LIGAND BINDING DATA

MS-DOS version 

The RBINDING.ZIP file should contain the following programs:

RBINDING.EXE     - the RBINDING program
RBINDING.TXT     - this documentation file    
RBG.EXE          - graphics program for RBINDING
RBG.BAS          - BASIC source code for RBG  
DEMO.RBD         - demo 'rough' data for RBINDING and RBG
DEMO.RBG         - demo fitted data as produced by RBINDING

DATE 30-11-1989

Binding analysis is performed according to the Linear
Subtraction method as described by Van Zoelen (1989)
Biochemical Journal 262, 549-556

These computer programs have been written by
Joop van Zoelen and Nico van Belzen
and should be considered as free scientific information

The program RBG is a graphical program designed for
Scatchard plots. Its use requires a graphical mode 
on your computer. The optimal scaling parameters
for this program may differ for individual computer
types. The RBG.BAS file is included in order to
make individual amendments possible!


RBINDING GENERAL STRATEGY

The RBINDING program analyses binding data of a radiolabeled
molecule to a specific binding site. The aim of the program
is to determine the number of binding sites and the binding
affinity. The program is split up into the following sections:

1:Entry of experimental binding data, including the possibility
  to correct these data and to store them on disk

2:Analysis of these binding data according to a single affinity
  receptor class, including the possibility to present these
  data according to a one receptor class Scatchard plot

3:Statistical analysis for the presence of multiple receptor
  classes

4:Analysis of the binding data according to a two receptor
  class model, including the possibility to present these
  data in a two receptor class Scatchard plot


section 1: RBINDING DATA INPUT

The following parameters are required for the analysis
a: incubation volume (ml)
b: specific activity of the labeled ligand (cpm/ng)
   (cpm is interexchangable with dpm)
c: molecular weight ligand
d: number of cells in the assay (in millions)
e: number of data points
f: (for each data point) conc. labeled ligand added (ng/ml),
   conc. unlabeled ligand added (ng/ml) and total binding
   of labeled ligand (cpm), uncorrected for any non-specific
   component

NOTE: In case all added concentrations are to be presented
in nanomolar instead of ng/ml, use for b: cpm/pmol and for
c: a molecular weight of 1000

NOTE: Data can also be expressed in other parameters related
to the cell number, for example in mg. protein. Enter the
amount of protein under d: ,and convert all obtained results
expressed in millions of cells to mg. of protein

The menu provides you with various options on data input
New data are to be entered through the keyboard (option#2)
Introduced data can be surveyed by option#8
And subsequently saved on the disk following option#3
(use a code name of not more than 8 characters)
Programs on disk can be recalled by option#1

NOTE: If new data are to be entered through the keyboard
the program will indicate in brackets the existing value
for this parameter in the memory. Only if the parameter
has to be changed, a new value should be introduced;
otherwise simply press ENTER.

NOTE: If corrections are to be made in the 5 general 
parameters indicated a-e on the previous screen, enter
the old program first from the disk, go back to the 
menu (M), and take option#2 for entering data from the
keyboard. Only the altered parameter has now to be 
introduced. The individual data points can be corrected
by the use of the options#5,6,7


section 2: SINGLE RECEPTOR CLASS ANALYSIS

After entering all data points analysis is started
following option#9

PROGRAM QUESTIONS (PQs)

PQ: Do you want to subtract a constant cpm value from
all data points ?
A: This can be used to subtract the counting background.
This option is not meant for subtracting estimated values
of non-specific binding. Only in the case of associated
radioactivity unrelated to the labeled ligand (for example
radioactive iodine dissociated from the labeled ligand),
a possible correction can be introduced here.


NOTE: ENTER skips Program Question !


SINGLE RECEPTOR CLASS ANALYSIS (continued)

PQ: Do you want to use another reference point?
A: The Linear Subtraction Method generally uses the first
data point (lowest concentration of only labeled ligand)
as reference point in the analysis, because the curve 
relating cpm bound to conc. unlabeled ligand is steepest
at low conc. unlabeled ligand. Theoretically, however, any
data point in the analysis can serve as reference point.
Only if there is serious doubt on the value of the first
data point, other reference points should be tested. If this
is done, only those reference points which yield "serious"
values for the parameters N, Kd and Ka (no negative values)
should be considered. The "serious" values for N,Kd and Ka
can subsequently be averaged for the various reference points.

NOTE that in case other reference points are used,
concentrations and cpm bound may take negative
values. The values for the binding parameters N, Kd and
Ka should however remain unaltered.



SINGLE RECEPTOR CLASS ANALYSIS (continued)

The program will now present the best fitting data for
the parameters N,Kd and Ka following the Linear Subtrac-
tion Method, based on a single receptor class analysis.

The program will then study the statistics of curve-fitting.
The implication of this analysis will be described in 
section 3: 

PQ: Do you want data expressed for single receptor 
Scatchard analysis?
A: (IMPORTANT) The Scatchard analysis uses the amount of
non-specific binding determined by the Linear Subtraction
Method, and therefore the two analyses are not independent.
Scatchard analysis is not essential, since all relevant
parameters (N,Kd and Ka) have already been established 
by the Linear Subtraction Analysis.




SINGLE RECEPTOR CLASS ANALYSIS (continued)
Single receptor Scatchard analysis (continued)

It is therefore recommended to use the obtained Scatchard
plot merely as an illustration plot. The data calculated
from the Scatchard plot should theoretically be similar
to those obtained from the Linear Subtraction Method, but
in practise they are not due to errors in the data points.
In general the data obtained using the Linear Subtraction 
Method should be considered as the best estimate for the
relevant binding parameters.

PQ: Do you want to test other values for Ka (the constant
for non-specific binding) ?
A: This is only relevant if multiple reference points have
been used, and a best (averaged) value for Ka has been ob-
tained. It also permits the possibility to find by trial-
and-error the value for Ka which gives the best linear fit
in a Scatchard plot. This approach is no longer relevant
when using the Linear Subtraction Method, and should be avoided
since data are forced into a single receptor class model.


section 3: STATISTICS OF NON-LINEARITY

The statistical analysis according to Weisberg determines
the probability for non-linearity of the Linear Subtrac-
tion Plot within a 90% probability interval. It should
be emphasized that non-linearity may be caused by the
presence of multiple receptor classes or as a result of
ligand binding cooperativity. In this latter case
binding cannot be described according to the equations
used in this program.

NOTE that the parameter Ti in the table denotes the
probability that individual data points can be con-
sidered as outliers (at Ti<-2 or >2 within 95% interval).
It is worth while to test if these data points have larger
standard errors than usual, indicating that one of
individual binding data making up the average value may
be significantly deviating from the expected behaviour.
It should be realized, however, that data points can behave
as outliers, because of the non-linearity of the Linear
Subtraction Plot, as an indication for the presence of
multiple receptor classes


section 4: TWO RECEPTOR CLASS ANALYSIS

The program will now test if deviations from linearity can
be explained by a two receptor class model. Based on a second
order correction, a first estimate for the high affinity
binding site is calculated. Subsequently a procedure is 
started to generate values for N and Kd of the low affinity
site (7*7 block), and to correct the Linear Subtraction Plot
for the presence of this low affinity site. For each combina-
tion of N and Kd of the low affinity site (7*7), a parameter
related to the deviation from linearity is calculated. Posi-
tive values of this deviation parameter are coupled to allowed
values for the N and Kd of the high affinity site (no negative
values). Using an iterative procedure the lowest positive 
deviation parameter is selected. In case this procedure will not
yield straightforward results, the program will suggest alter-
native options.
The data from the two receptor analysis can be plotted again in
a Scatchard plot, using the values for non-specific binding
obtained by the above method. The same limitations for the
validity of this Scatchard analysis apply here as for the
single receptor class Scatchard plot.


FOR SUGGESTIONS AND COMMENTS APPLY TO
JOOP VAN ZOELEN
DEPT. OF CELL BIOLOGY
UNIVERSITY OF NIJMEGEN
TOERNOOIVELD
6525 ED NIJMEGEN
THE NETHERLANDS

PHONE (31)-80-612707
FAX   (31)-80-553450

GOOD LUCK!!!


This program is also available for the Atari ST.
Contact   Nico van Belzen
          Dept. Pathology
          P.O.Box 1738
          3000 DR Rotterdam
          The Netherlands
e-mail: belzen@pa1.fgg.eur.nl
