Towards a semi-automatic functional annotation tool based on decision tree techniques

MLSB'07

Results : measures

n: number of proteins to be annotated
n_p: number of proteins with at least one prediction
n_p⁺: number of correct predictions
n_p^*: number of missing predictions
n_p^-: number of incorrect predictions

Multilabel - Tilde : curve of results obtained with a combination of Multilabel and Tilde systems ( if Multilabel predict nothing for a protein, we consider Tilde prediction for this protein)
Tilde - Multilabel : curve of results obtained with a combination of Tilde and Multilabel systems ( if Tilde predict nothing for a protein, we consider Multilabel prediction for this protein)
Multilabel : curve of results obtained with Multilabel system
Tilde : curve of results obtained with Tilde system

Hierarchical Precision: hP=n_p⁺ / (n_p⁺ +n_p^- )

Hierarchical Recall: hR=n_p⁺ / (n_p⁺ +n_p^* )

Hierarchical Fscore: hF_b=(b²+1).hP.hR /(b².hP+hR )

Menu MLSB07