Comparison of false-discovery rates of various decoy databases.

The FDR must be estimated using the correction factor proposed by Elias and Gygi or that by Kim et al. when (pseudo) shuffle decoy databases are used.