I am comparing the effect of tuberculosis on human genes. I am comparing it on Asian, European and Africans. I have 20 Asian samples, 10 European samples and 4 African samples (i.e the sample size is not in equal proportion). What would be the best statistical model to see how that comparison could be made to see changes in human genes by tuberculosis on different ethnicity?
Asked
Active
Viewed 172 times
0
-
What kind of data do you have? What kind of analysis are you planning? do you mean cases or observations by "samples", or is not how many cases do you have and how big is the discrepancy between sample sizes? – Tim Feb 16 '16 at 08:09
-
@Tim These are all TB cases. I do have unaffected controls to use for comparative analysis, but not sure how that is going to help for statistical analysis. If there were equal number of samples, I could have used t-test to check what genes are affected significantly, but the number of cases are disproportionate, I don't know how exactly I can show statistical significance. For example, 'these many Asians have this gene affected significantly, so is associated with Asian individuals and not with others..' – MAPK Feb 16 '16 at 08:19
-
See e.g. https://stats.stackexchange.com/questions/63856/power-of-the-t-test-under-unequal-sample-sizes – Tim Feb 16 '16 at 08:30