lsqcurvefit for biological replicates (three y values for one x value)

Question

Neha 2015-4-7

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/196555-lsqcurvefit-for-biological-replicates-three-y-values-for-one-x-value

编辑： Matt J 2015-4-7

在 MATLAB Online 中打开

Hi

I have a bunch of biological data that I am fitting to a model which looks like:

y=kcat*E.*x .*ATP./(KdATP*KmG+ KmG*ATP + KmA*x + x.*ATP.*(1+(x./KsiGlcn)));

Meausered values are: - x, E, ATP, KdATP, y

Parameters to be determined by fitting: - kcat, KmG, KmA, KsiGlcn

My data looks like:

experiment 1:
x E y ATP
0.14  6.65  0.12  1.00
0.29  6.60  0.21  0.98
1.85  6.08  0.30  0.79
1.98  6.03  0.30  0.77
experiment 2
x E y ATP
0.14  6.65  0.12  1.00
0.29  6.60  0.18  0.98
1.85  6.08  0.23  0.84
1.98  6.03  0.23  0.82
experiment 3
x E y ATP
0.14  6.65  0.08  1.00
0.29  6.60  0.11  0.99
1.85  6.08  0.14  0.89
1.98  6.03  0.14  0.89

So, there are three replicates (same experiment repeated thrice) which gives me three different y values for one x value. All the replicates are equally weighted. How should I fit this to my model?

I was using lsqcurvefit and it works well for a single experiment but when I combine the replicates I get a predicted y value for each x which is different. I just want one fitted value for each x. I hope I am explaining myself clearly.

Thanks in advance for any help.