See, i do understand that RI is for bulk shift accommodation when the running buffer and sample buffers are mismatched. but my question is when I chose a 1:1 model of curved fitting what criteria do i also look into to set RI=0 ? or i use the default setting. ? My sample buffer and running buffer is the same...
when should we set RI=0 for a 1:1 fit?
15 Apr 2015 19:38 #2
In general the RI is fitted along with the other parameters. Fitted RI should be low and proprtional to the analyte concentration.
I my opinion the RI is often calculated too large.
I always start with RI=0 to determine the Rmax, ka and kd. Then I add RI-fitting and start the fitting with the values of the previous fitting.
Lets clear something up.
If I use RI=0 I mean that I keep RI as a constant on zero. Thus assuming there is no RI effect. Ergo there can be no RI value other than zero.
If you set RI=0 and the fit it locally (must be local!) then the progam wil fit a value starting with zero. The default fittings use often a starting value of 10 or 50. When fitting clean sensorgram the answers should be comparable regardless the starting value (unless really extreme).
So I normally start fitting without RI --> RI=0 and a constant, Then after I have the ka, kd and Rmax, I fill in these values and set RI=0 and fit locally.
Maybe you can show the two fittings with values (or otherwise send a message via the contact form).
here i paste two sensograms... one with RI=0 and one with RI= fitted..
I have fcgRIIIa immobilised at 250 RU
there are 8 conc. of the drug product
Query is, I see when I use RI=0 the fitting is bad in 1:1 model and when I use RI= fitted i see a better fit... but i don't know if i could use RI=fitted or when I should use RI= 0...