I have about 50 accession and their origins are divided in 2 centuries (20 accession from American, 30 accession from European) and there are 10,000 gene for each accession so the dataframe looks like this ...

Now, I have to check whether there is statically valid difference between accession based on their origins. Basically, I want to compare American vs. European and get the genes which shows difference (pvalue <0.05)

I tried the ANOVA(lme4) or ttest(by using the mean values of America n and European group...) but have not decided or found the mort proper way or good way to figure out this. Do you have any recommendation or advice how to deal with this data?