Hello everyone.
I work with drug development, and I have been trying to apply some basic statistics concepts in my analysis. I would like to understand better what is a sample and what is the population, in the context of the IC50 value of a compound. Suppose I have some solution of a compound, and in 3 different days I evaluate the activity of this solution against a parasite. To the better of my knowledge, these three values would be considered as parts of a single sample (which would be described by its mean and standard deviation, for example). Suposing this is right, can I say that if I tested another solution (produced independently) of this compound, this new data would be of a second sample? What then would be considered as the population of IC50 values of this compound?
I hope I can reach someone with good statistics background and drug development knowledge. Thanks in advance
In the statistical modelling of biological variability and sample to population inference the investigator must distiguish technical variability and biological variability. If you are not interested in technical variability and are confident the second solution is the same as the first then the further IC50 measures could be added to the first sample and the whole sample used to statistically estimate the population mean IC 50 ( the mean value of a hypothetical infinite population of ic50 values) .
if you want to test that the two solutions do give the same IC50 you would need to treat the data from the two solutions as different samples and do a significance test of their difference. If they are not statistically significantly different you could then pool all the data to estimate ic50..
James Leigh, thanks for your brief and enlightening answer! That has certainly clarified some points to me. Do you have any reading suggestion on this specific topic? I have already read some statistics material, but I find most of them hard to translate to the drug development context. Thanks in advance
I think you have answered or solved the problem yourself. The first scenario where you are repeating your assays using the same solution of your compound from one pot will be replicates and therefore the mean will stand for one assay (a sample). On another day, when you prepare another pot of solution and do replicate assays, that will also be another sample. So values from assays performed from solutions of the same compound prepared on different days will give you the number of samples for calculating your IC50.
- Badges
- Science topic
- Similar topics
- Chemistry