Estimating the Confidence of Peptide Identifications without Decoy Databases