Email Post: Predicting mostly disordered proteins by using structure-unknown protein data