when doing genetic analysis, why do people not considering resampling methods, even the sample size is not big?

for example, I want to compare gene expression levels on case vs. control groups. if I take a subset of each group, and run differential expression analysis, i will get a list of differentially expressed genes (DEGs). and if I take another subset and do the exactly same analysis, i will get another list of DEGs, which wouldn't be same as the previous one. How do we think of this variance?

In statistics, people do resampling/ bootstrapping to deal with sample size issue. why we are not doing such thing and find a list of DEGs that more frequently showed up as the "true DEGs"?

Thank you so much!