using regex for this doesnt make any sense... just do
columns_with_2point5 = [c for c in df.columns if "2.5" in c]
only_cool_cols = df[['xvalues'] + columns_with_2point5]
dont overcomplicate it ...
if you dont need the first column you can just use filter with like instead of using one of the regex solutions (see first comment from @BeRT2me)