I have read through the documentation of corr.test and just am confused a little so would love some clarification. Does corr.test in r find the correlation between each column of two matrices and does adjusted p values based on the sample size of the columns being correlated? The reason I am asking this is because I performed corr.test between the columns of two matrices and got results. However, when I added more columns my numerical metadata matrix with still the same number of rows (samples) as my expression matrix, I get different p adjusted values for columns that were present in the first correlation. Should they not stay the same and just the new columns that are added in the second metadata matrix be just correlated by themselves with columns in the expression matrix. If anyone could please clarify, I would much appreciate this.