I think you mean the relative response after ligand immobilization. The relative response from the report point table is the response after immobilization (see
; point 7 minus point 3). Thus the amount of ligand that is immobilized on the sensor chip in RU. Theoretically the Rmax can be calculated from the amount of immobilized ligand and used analyte as you show in the inset of the figure you posted. The point is, this will only hold for a 1:1 interaction where all the ligand molecules are functional and accessible. All immobilizations will result is a fraction of ligand that is not functional anymore and the Rmax will be lower.
Also in the case of biotinylation of the ligand there is a possibility that a fraction is biotinylated on the functional binding site.
The next thing is that you use an antibody as an analyte. Choosing the bivalent model is a good thing but there are some problems that are difficult to resolve. An antibody can be bound to the surface by one or two arms and still give the same response. If all antibodies bind by two arms the theoretical Rmax should be divided by 2 (or set the valency at 0.5). If all antibodies bind by one arm the Rmax can be reached. The actual calculated Rmax during fitting will be probably somewhere in between and depends on the actual ligand density of the sensor chip.
I think you can convert the ka2 to M-1s-1 because I think that the SA chip is a modified CM5 chip. However, be careful with ka2 since it is (probably) dependent on the ligand density.
we see a significant difference between two groups. Treament gives higher RelResp mean vs placebo
But the Rmax reported after curve fitting the placebo Rmax is significant over the treatment which is the opposite
Not sure what you want to say/ask.
From the figure you posted the numbers are hard to read. What you can do is to simulate the curves with different analyte concentrations (maybe inject for two minutes to get some steady state) to see how the theoretical curves should look like.