| Author |
 Topic  |
|
|
sakthi
USA
Posts |
 Posted - 06/18/2004 : 1:21:44 PM
|
| hi, I am interested in calculating the hill coefficient for calcium dependent ATPase activity. I added values A (pCa) and B (ATPase activity). Then i choosed the scatter and then i choosed sigmoidal fit. I could then get values for Vmax and EC50, but not hill coefficient. I am blind in this program. I need a help from you to choose the exact result of hill coefficient. Or we have to do it manually! I don't know. Please help me. Thanks a lot, Sakthi |
|
|
easwar
USA
1965 Posts |
Posted - 06/18/2004 : 3:39:48 PM
|
Hi Sakthi,
I presume you performed the fit by using the menu item Analysis->Fit Sigmoidal. This menu option uses the Boltzmann equation.
I suggest you use the Sigmoidal Fit Tool, which can be brought up using the "Tools->Sigmoidal Fit" menu item. On this tool, in the Settings tab, click the Dose Response radio button in the "Logged Data Function" group and then go to Operations tab and click the Fit button. The fit results, resported in the Results Log window, will show a Slope parameter, which is the Hill Slope, and I presume this is what you are looking for?
Now, for advanced fitting, one can directly go to the NLSF Advanced Fitting tool, which can be brought up using the "Analysis->Nonlinear Curve Fit" menu item. All nonlinear fitting performed in Origin is based on what this tool can do. In this tool, you have finer control over things, such as selecting functions from various categories (Dose Response function is listed under Pharmacology category), and then see a preview of the equation, preview of typical curve shape, and control the fit by fixing parameters, setting bounds etc.
Easwar
OriginLab
|
 |
|
|
sakthi
USA
Posts |
Posted - 06/18/2004 : 4:34:49 PM
|
Hi EASWAR, Thanks for your kind response. I also got a response from your Technical person as
REPLY: Using Origin you can do a Least Squares fit to the Hill Equation. With
X&Y data in work sheet just do a CTRL+Y to bring up NLSF dialog. You need to
first select the Hill function from the the pre-defined functions. From menu:
Function/Select and select category Growth/Sigmoidal to select Hill Equation.
Then select Action/Data from menu to assign data, and Action/Fit to do the
fitting. Click on 100 Iter to do the Fit to the Hill function. If guesses are
good estimates the calculation should converge to correct value. This is a
brief description. You may need to consult the Help files for more details or
hit F1 key when dialog box is active to get information on input parameters.
Better still you could first review the project NLSF Built In Func.OPJ which
has a worked out example of using the NLSF function.
If you have further questions/concerns then contact us.
Fred Zimnoch
Technical Support Engineer |
|
|
|
sakthi
USA
Posts |
Posted - 06/18/2004 : 4:40:08 PM
|
Hi EASWAR,
For example,
Vmax - maximal value
k - this is EC50
n - what is n?
Which value I should select as a Hill coefficient.
For example:
pCa MgATPase activity
A (X) B(Y)
8 0
7 4.7
6.75 32.1
6.25 104.8
6.0 172
5.75 181
5.5 186
5 193
For which I got:
Vmax: 181.12
k - 6.28
n - -44.54
Now my question is which is the value for Hill equation or Hill coefficient. I assume that you called it as "y"
On the left lower window, I got lot of values including y
Chi 2/DoF R 2
--------------------------
18.56996 0.99
--------------------------
Which one is the value for Hill equation.
I would appreciate if you could send some more suggestions ASAP.
We have very old version of Origin. Thus i am trying with the Demo 7.5. Do you have special discount to buy Origin 7.5 for University of Cincinnati.
With best regards, Sakthivel, Cincinnati
|
|
|
|
sakthi
USA
Posts |
Posted - 06/18/2004 : 4:50:50 PM
|
Hi EASWAR,
What i did was, i entered these above values,
then select "Plot/Scatter" - there define X and Y values
and then select "Tools/Sigmoidal Fit"
There I now (after your e-mail) select "settings/dose response" and then again select "Operation/Fit"
As a result,
I got values at the left lower window
Lower
Upper
LogEC50
Slope (you mean this slope)?
But Fred refers different one where i could not find the same slope, i could see only "n"
THANKS YOU VERY MUCH FOR SPENDING TIME FOR ME. I need to calculate a lot for my paper.
Sakthi |
|
|
|
easwar
USA
1965 Posts |
Posted - 06/18/2004 : 10:25:15 PM
|
Hi Sakthi,
You first need to determine, by looking at the expressions of various fitting functions available in Origin, and/or by looking at some text book, as to what function you should be using.
I was suggesting Dose Response Function, which is:
y=A1 + (A2 - A1) / (1 + 10^( (logx0-x)*p) )
where x0 is EC50. I see from searching the web that the parameter p here is referred to as Hill Slope, or also as Hill Coefficient.
Fred suggested the Hill function, which is:
y=Vmax * x^n / (k^n + x^n)
I see from searching the web that n here relates to the Hill Coefficient, but not clear to me - could not find a page that clearly describes what Hill Coefficient is.
The bottom line is that Origin has a collection of fitting functions that may not have explicit parameter names, but the function form may be exactly what you are looking for. So you first need to know what function you need to be fitting with to see if it is available in the Origin built-in categories.
If the function you want to fit with is not available in the built-in category, then it is very easy to create a user-defined function in Origin, with parameter names of your choice, and then perform fitting.
If you know the exact expression you need to use, and/or the exact definition of the Hill Coefficient, please post again, or contact tech support for further help.
As for purchasing Origin 7.5 and discout information, please contact our sales personnel. We do offer academic discounts. I am not sure if there is a special arrangement with Univ. of Cinicinnati - the Sales group will know that.
Easwar
OriginLab
Edited by - easwar on 06/18/2004 10:27:47 PM |
|
|
|
sakthi
USA
Posts |
Posted - 06/18/2004 : 10:41:01 PM
|
| Hi EASWAR, THANKS A LOT FOR YOUR PROMPT RESPONSE. I WILL COME AGAIN TOMORROW MORNING WITH WHAT EXACTLY I NEED TO GET AND REFER YOU SOME PAPER. If you send your e-mail ID to my e-mail address, i will send you a copy of those papers. My e-mail ID <sakthi@cchmc.org>...Good night. Sakthi |
|
|
|
sakthi
USA
Posts |
Posted - 06/18/2004 : 10:44:45 PM
|
Hi,
I placed an order to buy the Origin 7.5. I assume that we have older version like 2.0 or something, since i don't see slope value on it. Bye! |
|
|
|
easwar
USA
1965 Posts |
Posted - 06/18/2004 : 10:55:59 PM
|
Hi Sakthi,
You can contact me via tech@originlab.com
Say "ATTN: Easwar" on the subject line, and at the top of the body mention that this is a follow-up on this forum thread.
Easwar
OriginLab
|
|
|
|
sakthi
USA
Posts |
Posted - 06/19/2004 : 11:11:25 AM
|
Hi EASWAR:
I sent an e-mail with a reference and my data.
Waiting for your reply,
Have a great weekend.
Thanks, Sakthi |
|
|
|
rlewis
Canada
253 Posts |
Posted - 06/19/2004 : 3:24:05 PM
|
The following which follows is the contents of an OriginC-driven NLSF function definition file which I use to fit enzyme activity data to the Hill equation. It should work in Origin 7.0 and Higher ...
[GENERAL INFORMATION]
Function Name=HILLCOEFF
Brief Description=
Function Source=N/A
Function Type=User-Defined
Function Form=Equations
Number Of Parameters=3
Number Of Independent Variables=1
Number Of Dependent Variables=1
Analytical Derivatives for User-Defined=Off
[FITTING PARAMETERS]
Naming Method=User-Defined
Names=Vmax,Khalf,N
Meanings=Maximal Y-value, X-value when Y=Vmax/2, Hill Exponent
Initial Values=100.0(V),3.5(V),1.8(V)
Lower Bounds=--(X, Off),--(X, Off),--(X, Off)
Upper Bounds=--(X, Off),--(X, Off),--(X, Off)
Number Of Significant Digits=6,6,6
[FORMULA]
double S=x^N;
y=Vmax*(S/(Khalf+S));
[CONSTRAINTS]
/*Enter general linear constraints here*/
[CONSTANTS]
[Parameters Initialization]
Vmax=max(y_data)*1.05;
Khalf=Curve_xfromY( &x_y_curve, Vmax/2.0 );
int i=0;
if(x_y_curve.GetSize()<10)
{
i=x_y_curve.GetSize();
}
else
{
i=10;
}
double sSum=0.0;
double lnK=ln(Khalf);
double sS;
int ni=0;
for (int ii=0;ii<i;ii++)
{
if(y_data[ii]==0)
{
continue;
}
sS=(lnK-ln((Vmax/y_data[ii])-1.0))/ln(Curve_xfromY(&x_y_curve, y_data[ii]));
sSum+=sS;
ni++;
}
N=sSum/=ni;
[INITIALIZATIONS]
/*Code to be executed before fitting, a good place for complicated initialization.*/
[AFTER FITTING]
/*Code to be executed after fitting, a good place for generating results.*/
[ON PARAM CHANGE]
/*Code to be executed when parameters change.*/
[INDEPENDENT VARIABLES]
x=
[DEPENDENT VARIABLES]
y=
[CONTROLS]
General Linear Constraints=Off
Initialization Scripts=Off
Scripts After Fitting=Off
Number Of Duplicates=N/A
Duplicate Offset=N/A
Duplicate Unit=N/A
Generate Curves After Fitting=Yes
Curve Point Spacing=Uniform on X-Axis Scale
Generate Peaks After Fitting=Yes
Generate Peaks During Fitting=Yes
Generate Peaks with Baseline=Yes
Paste Parameters to Plot After Fitting=Yes
Paste Parameters to Notes Window After Fitting=Yes
Generate Residuals After Fitting=No
Keep Parameters=No
Enable Parameters Initialization=0
Compile On Param Change Script=1
[COMPILE FUNCTION]
Compile=1
Compile Parameters Initialization=1
On Param Change Scripts Enabled=0
[ORIGIN C FUNCTION HEADER]
#pragma warning(error : 15618)
#include <stdio.h>
#include <data.h>
#include <math.h>
#include <utilities.h>
// Add your special include files here.
// For example, if you want to fit with functions from the NAG library,
// add the header file for the NAG functions here.
// Add code here for other Origin C functions that you want to define in this file,
// and access in your fitting function.
// You can access C functions defined in other files, if those files are loaded and compiled
// in your workspace, and the functions have been prototyped in a header file that you have
// included above.
// You can access NLSF object methods and properties directly in your function code.
// You should follow C-language syntax in defining your function.
// For instance, if your parameter name is P1, you cannot use p1 in your function code.
// When using fractions, remember that integer division such as 1/2 is equal to 0, and not 0.5
// Use 0.5 or 1/2.0 to get the correct value.
// For more information and examples, please refer to the "User-Defined Fitting Function"
// section of the Origin Help file.
[ORIGIN C PARAMETER INITIALIZATION HEADER]
#include <stdio.h>
#include <data.h>
#include <math.h>
#include <utilities.h>
// Add your special include files here.
// For example, if you want to use functions from the NAG library,
// add the header file for the NAG functions here.
// Add code here for other Origin C functions that you want to define in this file,
// and access in your parameter initialization.
// You can access C functions defined in other files, if those files are loaded and compiled
// in your workspace, and the functions have been prototyped in a header file that you have
// included above.
// You can access NLSF object methods and properties directly in your function code.
// You should follow C-language syntax in defining your function.
// For instance, if your parameter name is P1, you cannot use p1 in your function code.
// When using fractions, remember that integer division such as 1/2 is equal to 0, and not 0.5
// Use 0.5 or 1/2.0 to get the correct value.
// For more information and examples, please refer to the "User-Defined Fitting Function"
// section of the Origin Help file.
|
|
|
|
sakthi
USA
Posts |
Posted - 06/20/2004 : 09:06:20 AM
|
Hi Rlewis,
Thanks you very much for your support and suggestions.
Sakthi |
|
|
|
sakthi
USA
Posts |
Posted - 06/20/2004 : 09:07:43 AM
|
Hi EASWAR,
Your name looks like you are from TN. I am from Madurai. Did you get my e-mail with attached file. Looking for your detail message. Thanks again. Sakthi |
|
|
|
kaschorr
Germany
Posts |
Posted - 08/09/2004 : 06:07:44 AM
|
hi, I am interested in calculating the hill slope for the inhibition of human neutrophil elastase release. I used "growth/sigmoidal" categorie and the "logistic" function to get the curve. I could get values for IC50 (x0), but I need also the value for hill splope. I am also blind in this program. I need a help from you to choose the exact result of hill slope. Thanks a lot, Karin Schorr
Edited by - kaschorr on 08/09/2004 06:09:22 AM |
|
|
|
easwar
USA
1965 Posts |
Posted - 08/10/2004 : 1:53:46 PM
|
Hi Karin,
I think you need to fit with the Dose Response function in the Pharmacology category instead.
This function is defined in Origin as:
A1 + ( A2 - A1 ) / ( 1 + 10^( log x0 - x) * p )
where x0 corresponds to the EC50 value, x values are log of concentration, and p here is the Hill Slope parameter.
Easwar
OriginLab
|
|
|
|
kaschorr
Germany
Posts |
Posted - 08/11/2004 : 06:37:01 AM
|
Hi Easwar,
thank you for the answer! Is EC50 value the x0 or logx0 (categorie pharmacology, function dose-response)? I compared EC50 values from dose-response and logistic (categorie growth-sigmoidal) curves and the values are similar, but the p value (hill slope) are tottaly different (0.096 and 2.011). It means that the p value from logistic curve is not the hill slope, or what?
Thank you very much
Karin |
|
|
|
easwar
USA
1965 Posts |
Posted - 08/11/2004 : 3:44:52 PM
|
Hi Karin,
The Dose Response function under Pharmacology category that I listed above assumes that your x data is log of concentration.
Then, the third fit parameter that is returned is Log x0, which corresponds to Log EC50. So to get the EC50 value, you need to compute
EC50 = 10^(Log x0)
From other pharmacology sites, I see that the Dose Response function is the one which clearly has the Hill Slope parameter, which in Origin is called as just "p" in the equation/expression. Also the sites mention that the Dose Response function is also referred to as the 4-parameter Logistic Function.
Is your x data in log form?
Easwar
OriginLab
|
|
|
| |
Topic |
|
|
|
Hill coefficient