Evaluation of WDSP in predicting unknown proteins.

aThe positive WD40 protein dataset has 1402 relatively confident WD40 proteins that were selected by using the query in UniProt database: family: “wd repeat” AND domain: “wd repeats” AND annotation:(type:repeat wd) AND database:(type:smart wd) AND database:(type:pfam wd).bThe negative dataset consist of 2496 non-WD40 all-β proteins from the SCOP database and 4669 non-WD40 all-α proteins, α+β proteins or α/β proteins. As for all-β proteins, the homologues are deleted if the sequence identities are larger than 50%. The homologues of all-α proteins, α+β proteins and α/β proteins are removed with the identity cutoff at 30%.