Reply to the letter by Dhillon and colleagues