Hi Fellow Researcher,
I understand the title isn't clear enough. Allow me to explain. I have two pairs of cell lines (pair 1: cell line A, cell line A with WT gene 1; pair 2: cell line A, cell lines A with mut gene 1). The mutated gene 1 is not functional. I want to test if the targets of gene 1 are only deregulated (or more significantly deregulated) in pair 1 as compared to pair 2. Is there a way to show it?
I am thinking paired t-test on log fold change, but then it wouldn't show the direction of deregulation. (Although In my case, most targets are upregulated.) Taking absolute value of the log fold change would change the normality of the data, although there are non-parametric test for that.
May I ask if the community has done experiment that is similar? Is there a more elegant approach? Would linear regression work? Appreciate any input in the discussion.
Kent
Probably I believe, a linear regression would work for this. Specifically, you could use a linear regression model to compare the changes of gene expression values between the two pairs of cell lines. The model would take into account the log fold change of each gene as the dependent variable, and the cell line (pair 1 or pair 2) as the independent variable. probably, you may be able to assess the effect of the cell line on the log fold change of each gene, and determine which pair of cell lines had the greatest effect on gene expression. Moreover, you could use a post-hoc test such as a Tukey HSD to compare the differences between the two pairs of cell lines. Another approach you could take would be to use a Wilcoxon Signed Rank test, which is a non-parametric test that can compare the differences in two related samples. This test would be appropriate if the data is not normally distributed.
Best Wishes,
Romit
- Badges
- Science method